Q1 2024 Palantir Technologies Inc Earnings Call

and you to our first quarter 2024 earnings call. We'll be discussing the results announced in our press release issued after the market closed and posted on our investor relations website. During the call, we will make statements regarding our business that may be considered forward-looking within applicable securities laws, including statements regarding our second quarter and fiscal 2024 results, management expectations for our future financial and operational performance, and other statements regarding our plans, prospects, and expectations.

Today's call, we'll be discussing the results announced in our press release issued after the market closed and posted on our Investor Relations website. During the call. We will make statements regarding our business that may be considered forward looking within the applicable securities laws, including statements regarding our second quarter and fiscal 2024 results management's expectations for our future financial and operational performance.

And other statements regarding our plans prospects and expectations. These statements are not promises or guarantees and are subject to risks and uncertainties, which could cause them to differ materially from actual results information concerning those risks is available in our earnings press release distributed after the market close today and in our SEC filings, we undertake no obligation to update forward looking.

These statements are not promises or guarantees and are subject to risks and uncertainties, which could cause them to differ materially from actual results. Information concerning those risks is available in our earnings press release distributed after the market closed today and in our SEC filings. We undertake no obligation to update forward-looking statements except as required by law. Furthermore, during the course of today's call, we will refer to certain adjusted financial measures. These non-GAAP financial measures should be considered in addition to, not as a substitute for, or in isolation from, GAAP measures.

Except as required by law further during the course of today's call. We will refer to certain adjusted financial measures. These non-GAAP financial measures should be considered in addition to not as a substitute for or in isolation from GAAP measures additional information about these non-GAAP measures, including reconciliation of non-GAAP to comparable GAAP measures is included in our press release.

Additional information about these non-GAAP measures, including reconciliation of non-GAAP to comparable GAAP measures, is included in our press release and investor presentation provided today. Our press release, investor presentation, and other earnings materials are available on our investor relations website at investors.palantir.com. During the course of the call, we will refer to various growth rates when discussing our business. These rates reflect year-over-year comparisons, unless otherwise stated.

An investor presentation provided today, our press release Investor presentation, and other earnings materials are available on our Investor Relations website at investors at <unk> Dot com over the course of the call we will refer to various growth rates when discussing our business. These rates reflect year over year comparisons unless otherwise stated joining me on today's call are Alex Karp, Chief Executive Officer.

Ryan: Joining me on today's call are Alex Karp, Chief Executive Officer, Shyam Sankar, Chief Technology Officer, Dave Glazer, Chief Financial Officer, and Ryan Taylor, Chief Revenue Officer and Chief Legal Officer. I'll now turn it over to Ryan to start the call. We started the year exceedingly strong, with revenue of $634 million, an increase of 21% year over year, driven by the momentum of AIP and our continued strong performance in the U.S. Our results also highlight the growing strength of our U.S. government business and our enduring mission critical

Sean Thank our Chief Technology Officer, Dave Glaser, Chief Financial Officer, and Ryan Taylor, Chief Revenue Officer, and Chief Legal Officer, I'll now turn it over to Ryan to start the call.

Ryan D. Taylor: The continued interest in AIP is loud and clear in the conversations I'm having. We've shared our plans to capture the market with AI, and our results show that our strategy is not only successful; it is excellent, while still early days.

We started the year exceedingly strong with revenue of $634 million, an increase of 21% year over year, driven by the momentum of AIP and our continued strong performance in U S commercial.

Ryan D. Taylor: Our results also highlight the growing strength of our U S government business and our enduring mission critical work.

Ryan D. Taylor: The continued interest in AIP is loud and clear and the conversations I'm, having across our customer base.

Ryan D. Taylor: We've shared our plans to capture the market with the IP.

Ryan D. Taylor: And our results show that our strategy is not only succeeding at is accelerating.

Ryan D. Taylor: While still early days, our focus is on building the foundations of our long term business.

Ryan D. Taylor: Our focus is on building the foundations of a long-term relationship. We intend to relentlessly continue landing new customers and subsequently expanding those, as our products gain traction and have meaningful impact within Not only are we increasing the volume of new customers, but I'm pleased with our ability to grow these new customers. With regard to landing.

Ryan D. Taylor: We intend to relentlessly continue landing new customers and subsequently expanding those engagements as our products gained traction and have meaningful impact within enterprises.

Ryan D. Taylor: Not only are we increasing the volume of new customers, but I am pleased with our ability to grow these new customers as well.

Ryan D. Taylor: With regards to landing new customers, we've sustained our high volume of boot camps with over 915 organizations participating to date to meet inbound demand.

Ryan D. Taylor: We've sustained our high volume, with over 915 organizations participating to date to meet inbound. We are also seeing substantial deal cycles. As one example, a leading utility company signed a seven-figure deal just five days after completing a contract. Another customer immediately signed a paid engagement after just one day of their multi-day, and then converted to a seven-figure deal three weeks later.

Ryan D. Taylor: We are also seeing substantial deal cycle compression.

Ryan D. Taylor: As one example, a leading utility company signed a seven figure deal just five days after completing a boot camp.

Ryan D. Taylor: Another customer immediately signed up paid engagement. After just one day of their multi day boot camp and then converted to a seven figure deal three weeks later.

Ryan D. Taylor: We expect the favorable unit economics and higher throughput to continue to accelerate our, U.S. commercial is where we're seeing the greatest growth. While Q1 is seasonally our slowest quarter, AIP adoption by new and existing customers helps drive notable growth in customer acquisition and revenue in our U.S. commercial business. In Q1, we added 41 net new customers in the U.S. Our customer count increased 69% year over year and 19% quarter over quarter compared to 8% quarter over quarter growth in Q1 2020.

Ryan D. Taylor: We expect the favorable unit economics, and higher throughput to continue to accelerate our business.

Ryan D. Taylor: U S commercial is where we're seeing the greatest transformation.

Ryan D. Taylor: While Q1 is seasonally our slowest quarter.

IP adoption by new and existing customers helped drive notable growth in customer acquisition and revenue in our U S commercial business.

Ryan D. Taylor: In Q1, we added 41 net new customers in U S commercial our customer count increased 69% year over year, and 19% quarter over quarter compared to 8% quarter over quarter growth in Q1 2023.

Ryan D. Taylor: Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech; New customers span a variety of AIP's application CMS, from the largest independent bottling company in the U.S. to a leading energy and infrastructure company in a multinational area. In Q1, our US commercial business had customers from 56 of the 74. As we're landing new customers, we're also seeing those customers expanding. Across my conversations with customers, I've seen the recurring theme of them asking me how they can do more faster with enterprise transformation.

Ryan D. Taylor: Excluding strategic investments our U S commercial revenue soared by 68% year over year, and 22% quarter over quarter.

Ryan D. Taylor: New customers span a variety of industries as aip's applications seem endless.

Ryan D. Taylor: From the largest independent bottling company in the U S to a leading energy infrastructure company and a multinational airline.

In Q1, our U S commercial business had customers from 56 of the 74 gigs industries.

As we're landing new customers were also seen those customers expanding their work with us.

Ryan D. Taylor: Across my conversations with customers I've seen the recurring theme of them asking me how they can do more faster with enterprise transformation driven by AIP.

Ryan D. Taylor: We're showing them how they can move their AI strategy beyond. Existing customers such as Lowe's, Cleveland Clinic, and General Mills, among others, are realizing the extensive possibilities of AIIP within their own enterprise and increasing their scope accordingly. Lowe's accelerated its transformation from a starting point of no AI to utilizing production level AI for over 1,000 customer services. Resulting in a 75% reduction in overdue. As one of its directors noted, quote, we achieved this in just four months and onboarded a thousand users within three weeks of rollout.

We're showing them how they can move their AI strategy beyond chat.

Ryan D. Taylor: Existing customers, such as Lowe's, Cleveland Clinic, and general Mills, among others are realizing the extensive possibilities of AIP within their own enterprises and increasing their scope accordingly.

Ryan D. Taylor: Loews accelerated its engagement from a starting point of no AI to utilizing production level AI for over 1000 customer service agents, resulting in a 75% reduction in overdue tasks as one of its directors noted quote we achieved this in just four months and on boarded 1000 users with.

Ryan D. Taylor: And three weeks of rollout.

Ryan D. Taylor: Cleveland Clinic committed to a 10-year expansion deal to deploy it more broadly across, and General Mills expanded the scope of its work further last quarter, as its Senior Director noted, quote, "We're saving on average about $14 million annually, and it's really only deployed to part of our network."

Ryan D. Taylor: Cleveland Clinic committed to a 10 year expansion deal to deploy more broadly across its hospitals.

Ryan D. Taylor: General Mills expanded the scope of its work further last quarter as its senior director noted quote we're saving on average about $14 million annually and it's really only deployed to part of our network as we speak.

Ryan D. Taylor: We're seeing rapid expansions within key customers. For example, a Fortune 500 industrial company signed a three-year expansion deal, which increased the annual revenue run rate of our work with them nearly fivefold compared to our initial engagement with them in 2020. A Fortune 100 retail company started a pilot in Q2 2023, expanded to a use case conversion in August, then expanded its work to a $12 million ACV enterprise engagement last year. These are just a few examples.

Ryan D. Taylor: We're seeing rapid expansions within key customer accounts for example, a fortune 500 Industrials company signed a three year expansion deal, which increased the annual revenue run rate of our work with them nearly fivefold compared to our initial engagement with them in 2022.

Ryan D. Taylor: A fortune 100 retail company started a pilot in Q2 2023 expanded to a use case conversion in August then expanded its work to a $12 million ACB enterprise engagement last quarter.

Ryan D. Taylor: These are just a few examples more and more customers are expanding their work with us due to AIP and the incredible traction our software has within their organizations.

Ryan D. Taylor: More and more customers are expanding their work with us due to AIP and the incredible traction our software has within their organization. Now, turning to our U.S. government. Our revenue continued to accelerate, increasing 8% quarter over quarter versus 3% quarter over quarter in Q4, with our products every day having a critical impact on the current world. We see continued demand for Mission Management and Positive Reception to First. Last quarter, we were honored the U.S. Army awarded Palantir over $178 million to be the sole prime to build a next generation target under the tape.

Ryan D. Taylor: Turning to our U S government business, our revenue continued to accelerate last quarter, increasing 8% quarter over quarter versus 3% quarter over quarter in Q4 with our products every day, having critical impact on current world events.

Ryan D. Taylor: We see continued demand for mission manager and positive reception to first breakfast.

Ryan D. Taylor: Last quarter, we were honored the U S army awarded pounds here over $178 million to be the sole prime contractor to build the next generation targeting node under the Titan program.

Ryan D. Taylor: This marks the first time a software company has won a prime for a hardware and exemplifies Palantir's, the software, opening the door to vast new. It is with immense reverence that we approach building and maturing our revolutionary capabilities for our world, in our international government. We are continuing to ramp up the critical work for delivery of the UK NHS Federated Data Platform, as well as providing defense capabilities to ally partners around the world. Looking at our business and its impact I'm invigorated about the new year. We have never been more convinced about AIP and its power, as well as our continued efforts supporting the most critical missions around the world. I'll now turn it over to you.

Ryan D. Taylor: This marks the first time that a software company has won a prime contract for a hardware system and exemplifies pound his position as the software prime opening the door to vast new opportunities.

Ryan D. Taylor: It is with immense reference that we approach building and maturing our revolutionary capabilities for our Warfighters.

Ryan D. Taylor: In our international government business, we are continuing to ramp up the critical work for delivery of the U K NHS Federated data platform as well as providing defense capabilities to Allied partners around the world.

Ryan D. Taylor: Looking at our business and its impact broadly I am invigorated about the year ahead.

Ryan D. Taylor: We have never had more conviction about AIP and the power of our software as well as our continued efforts supporting the most critical missions around the globe.

Ryan D. Taylor: I'll now turn it over to Sean.

Speaker Change: Thanks, Ryan. The clear signal from AIP bootcamps is that AI is for builders. So many anecdotes and quotes from customers, all reinforcing the same point, they are getting more done in a day or two in AIP than in a quarter or two without. We have pioneered the approach to getting beyond chat and unlocking the value of LLMs in the enterprise, taking inherently unstructured inputs that are flying around the enterprise, be it emails, slacks, PDFs, images, comments, audio, and turning them into structured actions and outputs, taking an email from a customer requesting a different product mix and turning it into an actual inventory allocation in the ERP system of record, taking a health claims denial and programmatically generating the documentation and supporting evidence from the clinical records and contracts, automating PNC insurance claims, and even in government using vision models to narrow candidate products driving foodborne outbreaks at, And we have started rolling out build with AIP, a series of developer and builder oriented tutorials and reference implementations that enable builders to ramp quickly on the primitives and power of AIP and practical examples that unlock every employee at every customer.

Sean: Thanks, Ryan the clear signal from AIP boot camps is that AI is for builders. So many anecdotes and quotes from customers all reinforcing the same point they are getting more done in a day or two in AIP and in a quarter or two without.

Sean: We have pioneered the approach to getting beyond chat and unlocking the value of L. O labs in the enterprise, taking inherently unstructured inputs that are flying around the enterprise be it emails slacks Pds images comments audio and turning them into structured actions and outputs take.

Sean: Taking an E mail from a customer requesting a different product mix and turning it into an actual inventory allocation and the ERP system of record.

Taking a health claims denial and programmatically generating the documentation and supporting evidence from the clinical records in contracts.

Sean: Automating P&C insurance claims processing and even in government using vision models to narrow candidate products driving food borne outbreaks at CDC.

Sean: And we have started rolling out build with AIP, a series of developer and builder oriented tutorials and reference implementations that enable builders to ramp quickly on the primitives and power of AIP and practical examples that unlock every employee at every customer.

Speaker Change: Our growth is being driven by the incredible dynamism of the U.S. commercial market, and we believe the U.S. government will follow. With this momentum, we have launched builder boot camps in the U.S. government. The U.S. Army recently issued a memo identifying two Palantir systems, AIDP and Army Vantage, as amongst the five total platforms approved for builders.

Sean: Our growth is being driven by the incredible dynamism of the U S commercial market and we believe the U S. Government will follow with this momentum we have launched builder boot camps in the U S government U S. Army recently issued a memo identifying to pile onto your systems AIB P and army vantage as amongst the five total platform's approved for builders.

Speaker Change: The U.S. Army's Artificial Intelligence Integration Center, AI2C, at Carnegie Mellon, leverages these platforms for half of their active projects and recently built an application for the 18th Airborne Corps with OSDQ. A DoD customer recently hosted a hackathon showing the value of the open joint all-domain command and control, or JADC2 SDK that we have been pioneering. One participant commented that nominating targets with Gaia Assist turns a six-hour workflow into 10 seconds.

Sean: The U S Army's artificial intelligence integration center AI to see at Carnegie Mellon Leverages These platforms, where half of their active projects and recently built an application for the 18th Airborne Corps with OS decay.

Sean: Our Dod customer recently hosted a hackathon showing the value of the open joint all domain command and control or <unk> SDK that we've been pioneering one participant commented nominating targets with Guy S. S turns a six hour workflow into 10 seconds.

Speaker Change: We continue to invest in Mission Manager, and we'll be extending it to the Edge with our Edge X infrastructure in the U.S. Now customers can use their cloud instance as an integrated development environment for edge platforms to centrally build, test, and continuously deploy and manage multi-vendor Big Ten Edge applications. It covers everything from streaming pipelines, OSDK-backed applications, native Gotham applications, and third-party

We continue to invest in mission manager and we'll be extending it to the edge with our edge ex infrastructure in U S. Government now customers can use their cloud instance, as an integrated development environment for edge platforms centrally build test and continuously deploy and manage multi vendor big 10 edge ecosystems. It covers everything from screaming pipeline.

Sean: <unk> OSD Cape backed applications native Gotham applications and third party apps.

Speaker Change: We are excited with our team of rockstar partners to deliver on the US Army's Titan program. This marks the first time a software company has won a hardware contract, firmly establishing the role of the software prime. We believe the core of this software target workbench will be critical in every cockpit, every vehicle, and every. Finally, I'd like to acknowledge the eye-watering work of our service members and our allies in meeting the moment defending against the massive Iranian attack on Israel. The Gotham investments in JADC2 performed excellently, and we're building even more even faster. I'll turn it over to Dave to take us through the numbers. Thanks, Shyam. Q1 with a strong start to the year.

Sean: We are excited with our team of Rockstar partners to deliver on the U S. Army has tightened program. This marks the first time a software company has won a hardware contract firmly establishing the role of the software Prime we believe the core of this software our target workbench will be critical in every cockpit every vehicle and every ship.

Sean: Finally, I'd like to acknowledge the eye watering work of our service members and our allies and meeting the moment defending against the massive Iranian attack on Israel, the Gotham investments and <unk> performed excellently and we're building even more even faster.

Sean: Turn it over to David to take us through the numbers.

David: Thanks, Sean.

Q1 was a strong start to year.

David A. Glazer: Revenue growth accelerated to 21% year over year in the first quarter, driven by momentum in AIP and our US commercial business, and a reacceleration in our US government. We delivered our sixth consecutive quarter of got profitability, generating a record 106 million of got net income in the first quarter. We also delivered our fifth consecutive quarter of gap operating profit, generating a record 81 million of gap operating income in the, Adjusted operating margin expanded to 36% in the first quarter, continuing to highlight the strong unit economics of our The revenue and profitability outperformance drove a three point sequential increase to a rule of 40 score from 54 in the fourth quarter of 2023 to 57 in the first quarter of, This was the third consecutive quarter of an expanding Rule of 40 score, turning to our global top-line results.

David: Revenue growth accelerated to 21% year over year in the first quarter driven by momentum in A&P and our U S commercial business and a reacceleration in our U S government business.

David: We delivered our sixth consecutive quarter of GAAP profitability generating a record $106 million of got net income in the first quarter.

David: We also delivered our fifth consecutive quarter of GAAP operating profit generating a record $81 million of GAAP operating income in the quarter.

David: Adjusted operating margin expanded to 36% in the first quarter continuing to highlight the strong unit economics of our business.

David: The revenue and profitability outperformance drove a three point sequential increase to our rule of 40 score from 54 in the fourth quarter of 2023 to <unk> 57 in the first quarter of 2024.

David: This was the third consecutive quarter from an expanding rule of 40 score.

Turning to our global top line results we.

David A. Glazer: We generated $634 million in revenue in the first quarter, up 21% year-over-year and 4% sequentially, exceeding the high end of our prior. Excluding the impact of revenue from strategic commercial, revenue grew 24% year-over-year and 4%.

David: We generated $634 million in revenue in the first quarter up 21% year over year, and 4% sequentially exceeding the high end of our prior guidance.

David: Excluding the impact of revenue from strategic commercial contracts first quarter revenue grew 24% year over year and 4% sequentially.

David A. Glazer: Customer count grew 42% year-over-year and 11% sequentially to 554, and revenue from our largest customers continues to be strong. First quarter trailing 12 month revenue from our top 20 customers increased 9% year over year to 55 million. Now moving to our commercial business, first quarter commercial revenue grew 27% year over year and 5% sequentially to $290,000.

David: Customer count grew 42% year over year, and 11% sequentially to 554 customers.

David: Revenue from our largest customers continues to expand first quarter trailing 12 months revenue from our top 20 customers increased 9% year over year to $55 million per customer.

David: Now moving to our commercial segment.

David: First quarter commercial revenue grew 27% year over year, and 5% sequentially to $299 million.

David A. Glazer: Excluding impact from strategic commercial, first quarter commercial revenue grew 36% year over year, and We had a very strong quarter of commercial bookings. First quarter commercial TCV booked was 505 million, representing 187% growth year over year. Our US commercial business continues to see unprecedented demand driven by momentum from AI. First quarter U.S. commercial revenue grew 40 percent year-over-year and 14 percent sequentially to $150 million, surpassing international commercial revenue for the first time. Excluding revenue from strategic commercial, first quarter U.S. commercial revenue grew 68% year-over-year and 22%.

David: The impact from strategic commercial contracts first quarter commercial revenue grew 36% year over year and 4% sequentially.

David: We had a very strong quarter of commercial bookings first quarter commercial TCP booked was $505 million, representing a 187% growth year over year.

David: Our U S. Commercial business continues to see unprecedented demand driven by momentum from AIP first quarter U S. Commercial revenue grew 40% year over year, and 14% sequentially to $150 million, surpassing international commercial revenue for the first time.

David: Excluding revenue from strategic commercial contracts first quarter U S. Commercial revenue grew 68% year over year and 22% sequentially.

VIP is driving both new customer conversions and existing customer expansions in the U S. In the first quarter, we booked $286 million of U S commercial <unk>, representing a 131% growth year over year total many dual value in our U S. Commercial business grew 74% year over year and 14% sequentially.

David A. Glazer: AIP is driving both new customer conversions and existing customer expansion. In the first quarter, we booked 286 million US commercial TCV, representing 131% growth year-over-year. Total remaining deal value in our U.S. commercial business grew 74% year-over-year and 14%. Our U.S. commercial customer count grew to 262 customers, reflecting 69% growth year over year and 19% growth. We generated $149 million in international commercial revenue in the first quarter, representing 16% growth year-over-year but a 3% sequential decline as a result of continued headwinds in Europe and the revenue catch-up in Q4 that we noted last. We continue to capitalize on targeted growth opportunities in Asia, the Middle East, and beyond. Revenue from strategic commercial contracts was $24 million.

David: Our U S commercial customer count grew to 262 customers, reflecting 69% growth year over year and 19% growth sequentially.

David: We generated $149 million in international commercial revenue in the first quarter, representing 16% growth year over year, but a 3% sequential decline as a result of continued headwinds in Europe and the revenue catch up in Q4 that we noted last quarter.

David: We continue to capitalize on targeted growth opportunities in Asia, the middle East and beyond.

David: Revenue from strategic commercial contracts was $24 million in the quarter.

David A. Glazer: We anticipate second quarter 2024 revenue from these customers to decline to between 7 to 9 million, compared to 19 million in the second quarter. We continue to anticipate 2024 revenue from these customers to be approximately 2%, shifting toward government. First Quarter Government Revenue increased 16% year-over-year and 3% sequentially to $335,000.

David: We anticipate second quarter 2020 for revenue from these customers to decline to between $7 million to $9 million compared to $19 million in the second quarter of 2023.

David: Continue to anticipate 2020 for revenue from these customers to be approximately 2% of full year revenue.

David: Shifting to our government segment first quarter government revenue grew 16% year over year, and 3% sequentially to $335 million.

David A. Glazer: First quarter U.S. government revenue grew 12% year-over-year and 8% sequentially to $250,000. As Ryan noted, we're excited to be the sole prime contractor under the Titan program and will continue pursuing other defense. We believe we're well positioned to see growth in our US government business over the course of 2020. First quarter international government revenue grew 33% year-over-year and declined 9% sequentially to $79 million as a result of the revenue catch-up in Q4 that we noted last quarter and continued headwinds.

David: First quarter U S government revenue grew 12% year over year, and 8% sequentially to $257 million.

David: As Ryan noted we're excited to be the sole prime contractor under the tightened program and we'll continue pursuing other defense opportunities. We believe we're well positioned to see growth in our U S government business over the course of 2024.

David: First quarter International government revenue grew 33% year over year and declined 9% sequentially to $79 million as a result of the revenue catch up in Q4 that we noted last quarter and continued headwinds in Europe.

David A. Glazer: First quarter TCV booked was $904 million, up 128% year-over-year. Net dollar retention was 111%, an increase of 300 basis points from last. The increase was driven both by expansions at existing customers and new customers acquired in Q1. As net dollar retention does not include revenue from new customers that are acquired in the past 12 months, it does not yet fully capture the acceleration and velocity in our U.S. commercial business. We ended the first quarter with $4.1 billion in total remaining deal value, an increase of 22% year-over-year and 6% sequentially, and $1.3 billion in remaining performance obligations, an increase of 39% year-over-year and 5%.

David: First quarter, <unk> booked was $904 million of 128% year over year.

David: Net dollar retention was 111% an increase of 300 basis points from last quarter the.

David: The increase was driven both by expansions at existing customers and new customers acquired in Q1 of last year as net dollar retention does not include revenue from new customers that are required in the past 12 months. It does not yet fully captured the acceleration and velocity in our U S commercial business over the past year.

David: We ended the first quarter with $4 1 billion in total remaining deal value an increase of 22% year over year, and 6% sequentially and $1 3 billion in remaining performance obligations and increase of 39% year over year and 5% sequentially.

David A. Glazer: As a reminder, RPO is primarily comprised of our commercial business, and it is not taking into account contracts with an initial term of less than 12 months or Contractual Obligations that Fall Beyond Termination for Convenience Clauses, both of which are common. Turning to Margin Innings. Adjusted gross margin, which excludes stock-based compensation expense, was 83% for the quarter.

David: As a reminder, <unk> is primarily comprised of our commercial business as it does not take into account contracts with an initial term of less than 12 months and contractual obligations that fall beyond termination for convenience causes both of which are common in most of our government business.

David: Turning to margin and expense adjusted gross margin, which excludes stock based compensation expense was 83% for the quarter.

David A. Glazer: Adjusted income for operations, which excludes stock-based compensation expense and related employer payroll taxes, was $226 million, representing an adjusted operating margin of 36% and marking the sixth consecutive quarter of expanding adjusted operating income. Q1 adjusted expense was 408 million, up 2% sequentially and 2% year-over-year. Turning to the balance of the year, given our conviction in the U.S. business, coupled with our margin expansion, we intend to boost investment and resources in the U.S., including an AIP and specific defense software.

David: Adjusted income from operations, which excludes stock based compensation expense and related employer payroll taxes was $226 million, representing an adjusted operating margin of 36% and marking the sixth consecutive quarter of expanding adjusted operating margins.

David: Q1, adjusted expense was $408 million up 2% sequentially and 2% year over year.

David: Turning to the balance of the year, given our conviction in the U S business, coupled with our margin expansion, we intend to boost investment in resources in the U S, including in AIP and specific defense opportunities.

David A. Glazer: We expect expenses to ramp up starting in Q2 through the back half of the year, and we remain focused on calibrating expense growth below revenue growth for the full year in order to continue delivering on our goals of sustained GAAP profitability and GAAP output. In the first quarter, we generated gap operating income of 81 million, representing a 13% margin, our fifth consecutive quarter of gap operating income, and fourth consecutive quarter of expanding gap operating. We generated first quarter GAAP net income of $106 million, representing a 17% margin, our sixth consecutive quarter of GAAP profit. First quarter adjusted earnings per share was $0.08, and gap earnings per share was $0.04.

David: While we expect expenses to ramp starting in Q2 through the back half of the year remained focused on calibrating expense growth below revenue growth for the full year in order to continue delivering on our goals of sustained GAAP profitability and got operating income.

David: In the first quarter, we generated GAAP operating income of $81 million, representing a 13% margin our fifth consecutive quarter of GAAP operating income and fourth consecutive quarter of expanding GAAP operating margins.

David: We generated first quarter GAAP net income of $106 million, representing a 17% margin our sixth consecutive quarter of GAAP profitability.

David: First quarter adjusted earnings per share was <unk> <unk> and GAAP earnings per share was <unk> <unk>.

David A. Glazer: Additionally, our combined revenue growth and adjusted operating margin accelerated to 57% in the first quarter, a three-point increase to our Rule of 40 score from the prior quarter. We will continue to strive to maintain this exceptional balance between the top and bottom lines. Turning to our cash flow, in the first quarter, we generated $130 million in cash from operations and $149 million in adjusted free cash flow, representing a margin of 20% and 23%, respectively.

David: Additionally, our combined revenue growth and adjusted operating margin accelerated to 57% in the first quarter, a three point increase to our rule of 40 score from the prior quarter. We will continue to strive to maintain his exceptional balance of top and bottom line performance.

David: Turning to our cash flow in the first quarter, we generated $130 million in cash from operations and $149 million and adjusted free cash flow representing margin of 20% and 23% respectively.

David A. Glazer: In Q1, we also repurchased approximately half a million shares as part of our share repurchase program. As of the end of the quarter, we're approximately 990 million remaining of the original. We ended the quarter with $3.9 billion in cash, cash equivalents, and short-term U.S. Treasury.

David: In Q1, we also repurchased approximately <unk> 5 million shares as part of our share repurchase program as we ended the quarter with approximately $990 million remaining of the original authorization.

David: We ended the quarter with $3 $9 billion in cash cash equivalents and short term U S. Treasury securities, we retain access to additional liquidity of up to $500 million through our revolving credit facility, which remains entirely undrawn.

David A. Glazer: We retain access to additional liquidity of up to $500 million through our evolving credit facility, which remains entirely. Now turning to the route. For Q2 2024, we expect revenue of between $649 and $653 million, and adjusted income from operations of between $209,000 and $213,000. For full year 2024, we're raising our revenue guide. 2.677 and 2.689.

David: Now turning to our outlook for Q2 2024, we expect revenue of between 649 and $653 million.

David: Adjusted income from operations of between 209 and $213 million for.

David: For full year 2024, we are raising our revenue guidance to between $2 677, and $2 $6 9 billion, we're raising our U S commercial revenue guidance to in excess of $661 million, representing a growth rate of at least 45%.

David: We are raising our adjusted income from operations guidance to between 868 and $880 million. We continue to expect adjusted free cash flow of between $800 million and $1 billion.

David: And we continue to expect GAAP operating income and net income in each quarter of this year.

David A. Glazer: We are raising our U.S. commercial revenue guidance to in excess of $661 million, representing a growth rate of at least 45%. Additionally, we are raising our adjusted income for operations guidance to between $868 and $880 million. We continue to expect a just and free cash flow of between $800 million and $1 billion, and we continue to expect GAAP operating income and net income in each quarter. With that, I'll turn it over to Alex for a few remarks, and then Ana will start off. Welcome to our Q1 earnings. I think it is fair to say we crushed Q1 in the U.S. We are on fire. You see this in our rule of seven, and rule of 40 being 47.

David: With that I'll turn it over to Alex for a few remarks, and then onno will kick off the Q&A.

Alexander C. Karp: You see it in the 68 percent growth in U.S. commercial, do apples to apples, following on 70% growth last quarter, or 71% if you do apples to apples, taking out SPAC revenue. You see it in the deal growth in the U.S., growing from 70 to 136 in a year. You see it in the general enthusiasm around our products, especially in commercial, but also in government, which has begun to re-accelerate. You see it in general in our customer growth in U.S. com, which grew 69 percent.

Welcome to our Q1 earnings.

David: Aye.

Alexander C. Karp: Think it is fair to say, we crushed Q1 in the U S.

Alexander C. Karp: Our on fire that you see this in our rule of seven of rule of 40 being 40 57, you see it in the 68% growth in U S. Commercial if you do apples to apples following on 770% growth last quarter or 71%, if you do apples to apples.

Taking out of spec revenue.

Alexander C. Karp: You see it in the deal.

Alexander C. Karp: <unk> growth in the U S.

Alexander C. Karp: Growing from 70 to $1 36 in a year.

Alexander C. Karp: You see it in the general enthusiasm around our products, especially in commercial but also in government, which has begun to reaccelerate.

Alexander C. Karp: You see it in general and our customer growth in U S Com, which grew 69%.

Alexander C. Karp: And again, we are growing these numbers, while maintaining a rule of 40 score of 57, which basically means we're doing the impossible. We are growing the company, while and while investing in our core IP our software development.

Alexander C. Karp: And again, we are growing these numbers while maintaining a rule of 40 score of 57, which basically means we're doing the impossible. We are growing the company while investing in our core IP or our software development. We are winning in the U.S. And that kind of leads to the question of why are we doing so well in the United States of America?

We are winning in the U S.

Alexander C. Karp: And then that kind of asset asset then leads to the question of why are we doing so well in the United States of America.

Alexander C. Karp: And the main reason is that we built software infrastructure that allows enterprises, both commercial and government, to move beyond chat, move beyond self-pleasuring, and actually produce things that are valuable. And whether this is tasking satellites in the commercial context, or changing margins, or changing American workers into Japanese engineers using our software platform and large language models, we believe, I believe, we're the only company in America, the only really relevant market, that will allow you to do useful things with large language models.

Alexander C. Karp: And the main reason is we built software infrastructure that allows enterprises, both commercial and government to move beyond chat move beyond self pleasuring to actually produce things that are valuable.

Alexander C. Karp: And whether this is tasking satellites in the commercial context or changing margins are changing American workers into Japanese engineers, using our software platform and large language models. You can we believe I believe we are the only company in America, the only really relevant market that will allow you to do.

Alexander C. Karp: Do use full things with large language models.

Alexander C. Karp: And that is what's generating 68% growth in US com. Again, with a nascent Salesforce, doing it in a way that is not kind of playbook specific, thin technology, it barely works, and you have a Salesforce but with 10 years of IP that presupposed LLMs before LLMs existed.

Alexander C. Karp: And that is what's generating 68% growth in U S. Com again with a nascent salesforce is doing it in a way that is not kind of playbook specific thin technology. It barely works and you have a salesforce, but with.

Alexander C. Karp: 10 years of IP that presupposed L Ams before llm's existed.

Alexander C. Karp: You see the same thing in the government; Shyam and others will talk about it, but our mission footprint, whether it's in Ukraine, Israel, or the United States government, is stunning. There is basically no conflict in the world where Western allies are involved and the battlefield is involved, and the stakes are life and death, where Palantir is not the first call, and our ability to do this going forward is going to be even stronger because of past investments and also the unity of our culture and the strength of our leadership, and, in general, 20 years of understanding how to do this.

Alexander C. Karp: You see the same thing in the government charm and others will talk about it but.

Alexander C. Karp: Our Michigan footprint, whether its in Ukraine, Israel or in the United States government is.

Alexander C. Karp: Stunning.

Alexander C. Karp: There is basically no conflict in the world that is not in it's not that we're western allies are involved in the battlefields involved and the stakes are life-and-death, where talented or is not the first call.

Alexander C. Karp: And our ability to do this going forward is going to be even stronger because of.

Alexander C. Karp: Of the.

Past investments and also the unity of our culture and the strength of our leadership and in general 20 years of understanding how to do this.

Alexander C. Karp: It is true that it is mystifying to people, including analysts, how could this possibly be working, but it is working. And this is the sixth quarter of profitability. The sixth quarter of profitability. I dare remind people that there was a time when no one thought we'd be profitable, when no one thought we could crack commercial, where revenue per person was called into question. That also grew roughly 26% year on year; revenue per person generated at Palantir.

Alexander C. Karp: It is true that it is mystifying to people, including analysts how could this possibly be working but it is working and this is the sixth quarter of profitability. The six quarter profit of that idea remind people that there was a time when no. One thought we would be profitable when no one thought we could crack commercial where Ah.

Alexander C. Karp: At where revenue per person was called into question that also grew roughly 26% year on year revenue per person generated appellants here.

Alexander C. Karp: We do have headwinds in Europe, 16% of our business is in Europe. Europe is gliding towards zero percent GDP growth over the next couple of years, that is a problem for us; there is no easy remedy for that. It is also the case, though, that U.S. comps surpass U.S. European business, and we see that as favorable. And you know, I generally think this is just Palantir's time. Now there are lots of questions about why we are so active in defending the values of the West.

Alexander C. Karp: We do have headwinds in Europe, 16% of our business in Europe Europe is guy is gliding towards zero.

Alexander C. Karp: Zero percent GDP growth over the next couple of years that is a problem for US there is no easy remedy for that.

Alexander C. Karp: He is also the case, though that U S comp surpassed U S European commercial and we see that as favorable.

Alexander C. Karp: And.

Alexander C. Karp: I in general think.

Alexander C. Karp: This is just powershares time know it.

Alexander C. Karp: Lots of questions about why we are so active in defending the values of the west.

Alexander C. Karp: Our belief that the West is a superior way to live and our ways of organizing around that are the reason why our products are transformative, the reason why we have the best people in the world, the reason why a Palantir degree, as it were, is much more valuable than an Ivy League degree, before the Ivy Leagues even embraced the thin, new, woke religion otherwise viewed as an intellectual cause, but in fact is a way of organizing things so that the greatest institutions of our time disappear and turn into discriminatory dysfunction.

Alexander C. Karp: The our belief that the worst is a superior way to live and our ways of organizing around that are the reason why our products are transformative. The reason why we have the best people in the world. The reason why a pound here degree as it were is much more valuable than an Ivy League degree before the Ivy League even embraced the thing.

Alexander C. Karp: The new woke religion, otherwise viewed as an intellectual cause but in fact is a way of organizing things. So that the greatest institutions of our time disappear and turn into just discriminatory dysfunction, powershares, a counterexample and I'm Super proud of the results.

Alexander C. Karp: Palantir is a counterexample, and I'm super proud of the results. We are going to continue to execute, especially in the U.S., and I'm very happy to have you on the call. Thanks, Alex. We'll now turn to questions. We received a few questions from our shareholders about AI. Matthew and Ryan asked, "How does your AI strategy differ?" Well, we're executing our strategy at an unrelenting pace here. If you go back to the launch of AIP, I discussed how we thought the models were going to commoditize, and that's really borne out. We see that underlined with LLAMA370B being released now.

Alexander C. Karp: We're going to continue to execute especially in the U S and I'm very happy to have you on the call.

Speaker Change: Thanks, Alex we'll now turn to questions. We've received a few questions from our shareholders about AI and competitiveness, Matthew and Ryan ask how does your AI strategy differ from your competitors.

Speaker Change: Well, we are executing our strategy and an unrelenting pace here. If you go back to the launch of AIP I discussed how we thought the models are going to commoditize and that's really borne out we see that underscored with Lama $3 70, being released now, but the real opportunity for us as Alex made mention of is that people are using LMS incorrectly in the enterprise.

Alexander C. Karp: But the real opportunity for us, as Alex made mention of, is that people are using LLMs incorrectly in the enterprise. As far as I can tell, we're really the only company to figure out how to help our customers get beyond chat, leveraging the investments that we've made in ontology, really harnessing this pattern of implementation where you're taking unstructured inputs and turning them into structured actions and outputs that drive economic value in the enterprise.

And as far as I can tell we're really the only company to figure it out how to help our customers get beyond chat leveraging the investments that we've made an ontology really harnessing this pattern of implementation, where you're taking unstructured inputs and turning them into structured actions and outputs that drive economic value in the enterprise now there is this.

Alexander C. Karp: Now, there's this thing that some companies have started saying where only 10% of my customers have data that's even AI-ready to begin with. I think that's completely ridiculous. Maybe they don't have anything to sell in the present moment, so they're trying to sell the past.

Speaker Change: Thing that some companies have started saying, we're only 10% of my customers have data, that's even AI ready to begin with I think that's completely wrong.

Speaker Change: Maybe they don't have something to sell in the present moment, so theyre trying to sell the past, but if that was right. How is it that in a single day bootcamp, we're able to add value on top of our customers' messy extent data ultimately software that works works and to the present moment I'm focused on helping enable builders in the context of the enterprise I made mention of the Dod.

Alexander C. Karp: But if that was right, how is it that in a single-day boot camp, we're able to add value on top of our customers' messy extant... ultimately, software that works. And at the present moment, I'm focused on helping enable builders in the context of the enterprise. I've made mention of the DoD hackathon, where a single user built an AIP logic function, surfaced in Gaia Assist, that took a targeting process down from six hours to 10 seconds.

Speaker Change: Hackathon, where a single user built in AIP logic function surfaced and Gaia assessed that took a targeting process down from six hours to 10 seconds.

Alexander C. Karp: More generally, in conjunction with Build with AIP, we are releasing a slew of tutorials, quick starts, and reference implementations. They're going to help turn every, at every customer, every user into a builder to unlock the potential of what they can harness on top of AIP. And that's the way ahead, and that's why we're still ahead. But I would say, I don't believe we have. So I don't believe in the U.S. commercial market; we have competition. I don't believe in the U.S. government market; we have competition.

Speaker Change: More generally in commercial with build with AIP, we are releasing a slew of tutorials quick starts and reference implementations that are going to help turn every at every customer every user into a builder to unlock the potential of what they can harness on top of the AIP.

Speaker Change: And Thats the way ahead, and that's why we're still ahead.

Speaker Change: I'd say.

Speaker Change: I don't believe we have competitors. So I don't believe in the U S. Commercial market. We have competition I don't believe in the U S government market. We have competition I don't I think that's the reason Ukraine and Israel bought our product we are differentiated because in order to actually make AI work you need an ontology no. One has an ontology to shops point you have a lot of.

Alexander C. Karp: I think that's the reason Ukraine and Israel bought our product. We are differentiated because in order to actually make AI work, you need an ontology. No one has an ontology.

Alexander C. Karp: To Shyam's point, you have a lot of people running around saying the data isn't ready. Of course, it's not ready because they don't have Foundry. If you have Foundry and the ontology, it is ready. If you have Foundry or ontology in Apollo, you can actually work at the edge. If you don't, you can't.

Speaker Change: People running around saying the data isn't ready of course, not ready because they don't have foundry foundry and.

Speaker Change: And the ontology. It is ready if you are foundry the anti.

Speaker Change: Our ontology and Apollo you can actually work at the edge. If you don't you can't and you know it's outside of America, they're still we would still have to convince people of this inside of America, we're not really convincing people that we have the only thing that works, we're showing up we're showing it working and we're saying this is what it costs and it's working really well, but I I currently I don't.

Alexander C. Karp: Outside of America, we would still have to convince people inside America that we have the only thing that works. We're showing up, we're showing it working. And we're saying this is what it costs. And it's working really well, but currently, I don't believe we have competition. We have a lot of people who are like the Palantir of Iowa or the Palantir of Harvard or the Palantir of Uruguay, but they're not Palantir.

Speaker Change: We have competition, we have a lot of people who are like the challenge here of Io or the power interior of Harvard or the talent here of Uruguay, but theyre not pounds here.

Alexander C. Karp: And it's going to take a long time to actually build what we have because you'd have to actually understand what is special about the software infrastructure of having the combination of Apollo, Foundry, and Ontology. And then you would have to build on top of it so that you can actually do these handoff functions with large language models, and luckily for us, there is not a consensus at the investor level, the VC level, or among analysts that we have the only thing that works because, that means there's just going to be very little investment in copying and doing, and investment in what we do outside of America would be hard to do anyway. Ecosystem, a term I don't particularly like, to get this done is in America.

Speaker Change: And it's going to take a long time to actually build what we have because you'd have to actually understand what is special about it.

Speaker Change: About about the software infrastructure of having the combination of Apollo foundry and the and the ontology and then you would have to build on top of it. So that you can actually do these handoff functions.

Speaker Change: With large language models and Luckily.

Speaker Change: Luckily for US there is not a consensus at the investor level V C level or among analysts that we have the only thing that works because and that means there's just going to be very little investor investment and copying and doing what we do and investment in what we do outside of America would be hard to do anyway, because the ecosystem of term I don't particularly like to get this.

Speaker Change: One is in America. So you would have to do this in Silicon Valley and Silicon Valley is focused on to Sean's point the wrong things.

Alexander C. Karp: So you would have to do this in Silicon Valley, and Silicon Valley, thank you both. Our next question is from Dan with Wedbush. Dan, please turn on your camera, and then you'll receive a prompt to unmute your line.

Speaker Change: Thank you bet. Our next question is from Dan with Wedbush, Dan. Please turn on your camera and then you'll receive a tiny airline.

Speaker Change: Thanks.

Dan: Thanks. So, good, great quarter. My question is, can you just talk about what conversion looks like from bootcamps? And maybe just double-click on a typical customer? What's now that you have more and more data points? What's that showing you about conversion from a bootcamp to actually signing a deal? Yeah, absolutely. I can talk through that.

So good great quarter.

Dan: My question is can you just talk about what conversion looks like from boot camps, and maybe just double click on a typical customer what now you have more and more data points, what's that showing you about conversion from a boot camp to actually signing a deal.

Speaker Change: So we announced bootcamps, you know, two quarters ago as a go-to-market motion, and we're seeing that play out, as you know, you see that in the results. So in one to five days with a bootcamp, we're able to do what used to take three months. And we're seeing, as I talked about, customers, shortly after bootcamp, sign seven-figure deals. And we're seeing, you know, the ability to be able to show them what they can do on the platform with real data much more quickly, and then monetize that.

Yeah, absolutely I can talk to you that's what we announced boot camps.

Dan: Two quarters ago as a go to market motion and we're seeing that play out as.

You see that in the results. So in one to five days with a boot camp, we're able to do what used to take three months and we're seeing as I talked about we're seeing customers.

Dan: Shortly after bootcamp signed seven figure deals.

Dan: And we're seeing the ability to be able to show them. What they can do on the platform with real data much more quickly and then have that monetization conversation much sooner and so then you see that in the results 69% growth in customers and U S. Commercial we closed 136 deals in Q1 this year compared to 70 deals.

Speaker Change: Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, year over year increase. And so we're seeing, you know, it's still early in the process, and we're seeing the results from the boot camps and seeing that they're working.

Dan: In Q1 last year and U S. Commercial that's a 94% year over year increase and so we're seeing it's still early in the process and we're seeing.

Dan: The results from the boot camps and seen that they are working in the monetization from the boot camps also have quite frankly, another massive advantage because de facto it sets a standard that it'll be very hard for any other company to meet so even if you don't buy a product you de facto of locked in an idea of what's possible and that means that at some point when you go try your own.

Speaker Change: We also have, quite frankly, another massive advantage because, de facto, it sets a standard that will be very hard for any other company to meet. So even if you don't buy our product, you de facto have locked in an idea of what's possible. And that means that at some point, when you try your own thing and it fails, you've seen, okay, well, I've seen the art of the possible, and therefore, I'm going to go buy it from Palantir at some point in time. The bootcamp motion is an early motion.

Dan: Thing and it fails you have seen okay, while I've seen the art of the possible and therefore I'm going to go buy it from pound here at some point in the future. The boot camp motion is an early motion and I don't think we would say we've cracked the sales motion in fact, I think our 68% or 94% growth on deals are.

Speaker Change: And I don't think we would say we've cracked the sales motion. In fact, I think our 68% or 94% growth on deals or 69% on customers, you have to look at this as we are in the very early days of figuring out how to actually get customers to buy our product. But it is really early days, and we're not flawlessly executing on our sales. And I would say I saw this when we built our anti-terror platform. One of the most important things we did was go out and educate the world about how we could fight terror and maintain civil liberties.

69% on customer.

Dan: You have to look at this as we are at the way early days of figuring out how to actually get customers to buy our product. We are good at educating customers on what is the art of the possible and then some portion of those customers buy it. So I'd expect as we get better and better at that our numbers will increase.

Dan: But it is really early days, it's not we're not flawlessly executing on our sales motion.

Dan: I would say I've seen I saw this when we built our antiterror platform.

Dan: One of the most important things we did is go out and educate the world about how you could fight terror and maintain civil liberties and most people, we actually offered our platform due to did not buy the product year, one but by year five date all bought it.

Speaker Change: And most people we actually offered our platform to did not buy the product year one, but by year five, they'd all bought it. And they bought it because they were like, okay, well, we see how you could integrate data across disparate data sets, maintain a security model, fight terrorism, and be in conformity with the law. And even if you say, Oh, I'm going to go try to do this myself. In fact, you can't, and then that sales motion happens.

Dan: And they bought it because they are like okay, well, we see how you could integrate data across disparate datasets maintain a security model via terrorism and being conforming with law and even if you say Oh I'm going to try to do it myself in fact, you can't.

Speaker Change: So the integral, the relevant integral for this is not just what you close in the quarter; it's like, it is kind of throwing a carpet on the whole market and saying, okay, well, this is the standard; you can do better than this by yourself. Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, And that worked very, very well. I mean, almost every country in the world, not every country, but almost every country in the world that could buy our core anti-terror product.

Dan: And then that sales motion so the integral.

Dan: Relevant integral for this is not just what you closed in quarter. It's like it is kind of throwing a carpet on the whole market and saying okay. Well. This is a standard you can do better than this by yourself build it yourself acquired from other people and if you can't come back to US we're here.

Dan: And that worked very very well I mean almost every.

Dan: Our country in the world not every country, but almost every country in the world that could buy our core antiterror product at some point it.

Speaker Change: Thank you. Our next question is from Mariana with Bank of America. Mariana, please turn on your camera, and then you'll receive a prompt.

Dan: Thank you. Our next question is from Mariano with Bank of America Marianna. Please turn on your camera and then you always see the prompt on your line.

Mariana Perez Mora: Good afternoon, everyone. So it's also in the line of implementation, but less about the customer because I think the demand you're seeing is unprecedented. And we can see that you're seeing that on both sides, like from the government side and the commercial side. What I'm really curious about is if you can give us some color around what is next in terms of actually making, tackling this demand, implementing the efforts, and supporting a way larger customer base. What is hiring?

Mariano: Sterno and everyone.

Mariano: So it's also in the line of implementation that less about the customer because.

Mariano: I think the demand you are seeing in some parts around that and we can see that you are seeing that both sides like from the government side on the commercial side, but I'm really curious about is if you can give us some color around what is next in terms of actually making tackling this demand and implementing efforts in supporting the way larger customer.

Mariano: Base, what is hiring wise partnerships what is snacks. There what is the most challenging from from your point of view.

Mariana Perez Mora: What are partnerships? What is next there, and what is the most challenging from your view? Well, I can certainly take that maybe Ryan wants some comments on the partnership side and the growth of that ecosystem. But on the product side, this is one of the reasons I'm most excited about building with AIP and the bootcamps that we're running after we have deals with customers. We're really igniting a movement within the customer in terms of unleashing their builders and tackling. There's been a shift.

Speaker Change: Well I can certainly take that maybe Ryan will have some comments on the partnership side and the growth of that ecosystem, but on the product side. This is one of the reasons I'm. Most excited about build with AIP and the boot camps that we're running after we have deals with customers, we're really igniting a movement within the customer in terms of unleashing their builders and tackling use cases theres been a shift like we.

Mariana Perez Mora: We still love to work on our customers' hardest problems and most important use cases to deliver crushing value very quickly. But concomitant with that, how do we enable every single one of their builders to get going? And I can think of the last, say, two quarters, our ability to come in, help them solve one problem, but ignite hundreds of use cases that they're able to do on their own. So really without needing additional partners, leveraging repeatable reference implementations, reference architectures, quick starts, tutorials that get them going.

Speaker Change: Still love to work on our customers' hardest problems and most important use cases to deliver a crushing value very quickly, but concomitant with that is how do we enable every single one of their builders to get going and I can think of the most recent say two quarters, our ability to come in there help them solve one problem, but ignite hundreds of use cases that they're able to do on their own.

Speaker Change: So really even without needing additional partners leveraging repeatable reference implementations reference architectures quick starts tutorials that get them going again and across different we have these for all these industries that we're working in for all of the functions within these customers and including in government itself.

Mariana Perez Mora: Again, and across different verticals, we have these for all the industries that we're working in, for all the functions within these customers, and for government. And on the partnership front, you see, you know, we announced our partnership with Oracle; we're seeing hyperscalers that are realizing that in order to drive compute, you need to move beyond chat, right? Organizations beyond their, you know, for their AI strategy beyond chat, we're seeing Partnerships across across We have, across our company for the first time, ongoing discussions with, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech and build on top of our product, at a higher margin. We are having discussions like this in Japan, across the U.S.

Speaker Change: And on the partnership front, you'll see we announced our partnership with Oracle, we're seeing hyperscale or is that I realize that in order to drive compute.

Speaker Change: You need to move beyond chat right you need a solution that's taking organizations beyond there.

Speaker Change: Our strategy beyond chat, we're seeing that happen there and we're expanding the partnerships across across the space with partners implementing at our customers as well.

Speaker Change: Across our company for the first time ongoing discussions with Hyperscale or is that are well known with others, where we're basically saying look we have the ideal product you have the ideal distribution you can build on top of our product and then capture your distribution at a higher margin.

Speaker Change: We are having discussions like this in Japan across U H G, but we have.

Mariana Perez Mora: But we have the efforts to provide our core infrastructure to defense tech startups, which are going very well simultaneously allow us to expand our core architecture, core software development into the DoD, but also make us less competitive with people, which helps us a lot. So a lot of what slows us down is people pretending they're, so finding ways to build partnerships so they don't compete, Xcel or HR. And in USCOM, a lot of what Shyam is saying.

Speaker Change: Our efforts to provide our core infrastructure to defense Tech startups, which are going very well simultaneously allow us to expand our core architecture <unk> software development into the D. O D. But also make us less competitive with people, which helps us a lot. It's a lot of what slows us down as people pretending they're competing with us so finding ways.

Speaker Change: To build partnerships with those people. So they don't compete accelerates our revenue and niche and U S. Com. It's it's a lot of what Sean was saying, it's a we are allowing customers to do what we had to do in the past, which is supply power to your software engineers that are just not that many pounds tyrion. So we can't scale, we have to allow other encourage and provide a platform where they can do it.

Mariana Perez Mora: It's, we are allowing customers to do what, apply Palantir Software. Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, and that's again why that you see a revenue growth of up to 26% compared to last year on per person revenue; we're making more money. Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, And so that has a lot of benefits for us, both alignment and quality of revenue. That's why we have 57 on the Rule of 40.

Speaker Change: And I would say what you see both on AIP and on foundry is the facto customers using our product are pretty much on their own.

Speaker Change: And Thats again, why that you see our revenue growth up to 26% of compared to last year on per person revenue, because we're making more money with the same number of people roughly so and that's precisely because of these things that we're doing.

Speaker Change: And so there's a lot of benefits for us both alignment and quality of revenue. That's why you have this that's why we have 57 on the rule of 40, that's why that's why our revenue is growing per person, that's why where we can get away with sales that are not perfect.

Mariana Perez Mora: That's why our revenue is growing per person, that's why we can get away with sales that are not perfect, and we're thinking a lot about you turning back to a question from our shareholders. A lot of people have noticed that Palantir is pretty outspoken when it comes to geopolitics.

Speaker Change: And we're thinking a lot about how to do that.

Speaker Change: Thank you.

Speaker Change: Turning back to a question from our shareholders a lot of people have noticed that <unk> pretty outspoken when it comes to geopolitics. How has this been received.

Mariana Perez Mora: Um, well, you know, it's a complicated question because I think internally, there have been people who disagreed. There have been Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, arguably more controversial. The fact that I take, you know, I think the central risk to Palantir and America and the world is a regressive way of thinking that is corrupting and corroding our institutions that call themselves progressive, but actually, and it's called "woke," but it's actually a form of a thin pagan religion.

Speaker Change: Well you know, it's a complicated question because I.

Speaker Change: I think internally I mean, there have been people have disagreed there have been people, who disagree to the point, where they left and this has gone on over the last 20 years when we Paul when we refused to walk away from special forces. The U S. Army. When we took over May then when we refused to stop working with homeland security.

Speaker Change: When we when.

Speaker Change: When we have discussions about which countries you work with and which countries we wouldnt.

Speaker Change: More lately more more tie in a more timely version when we supported Ukraine was less controversial when we when we when we sold Israel our software that was.

Speaker Change: Arguably more controversial.

Speaker Change: The fact that we that I take I think the central risk to balance here and America and the world is.

Speaker Change: A regressive way of thinking that is corrupting and corroding are institutions that calls itself progressive but actual and is called woke, but is actually a form of a thin pagan religion that is a real danger to our society and it is a real danger to pound share. If we allow if we don't discuss these things.

Mariana Perez Mora: That is a real danger to our society, and it is a real danger to Palantir if we allow it, if we don't. The reason we have by far the best product offering in the world is because we have by far the best alignment around how to build software, what it means to build software, and full alignment with our customers. Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech. I mean, honestly, everyone at this table has a lot of money.

Speaker Change: The reason, we have by far the best product offering in the world is because we have by far the best alignment around how to build software what it means to build software for alignment with our customers a view that some of the western way of living is superior and therefore, it should be supported by the best products people at this take.

<unk> been in the trenches for over a decade, each why do they stay why do they fight why do we come and do these costs why are we fighting for individual investors why do why why are we ask why do we actually care about revenue growth in quarter.

Speaker Change: Honestly, everyone. At this table is a lot of money.

Mariana Perez Mora: It's like we care, and we fight for these things because we believe we are fighting for a stronger, better... West discriminatory, wealthier, more open, and better society by providing the friends of the West, U.S. industry, U.S. government, our allies, with far superior products.

Speaker Change: It's like we care and we fight for these things because we believe we are fighting for a stronger better west discriminatory wealthier more open and better society by providing the friends of the West U S industry U S government, our allies with by far superior products and that's <unk>.

Mariana Perez Mora: And that's, by the way, how we, you would be, and that's how we align with our customers. By the way, I have customers who disagree with me, but they sure as hell know I'm telling the truth about what we believe. And probably more importantly, they know we're not sophists. They know we believe them.

Speaker Change: By the way how we are you would be and that's how we align with our customers by the way I have customers, who disagree with me, but they sure as Hell no I'm, telling the truth about what we believe and probably more importantly, they know we're not soft surface. They know we believe in things.

Speaker Change: And that's why we fight for this culture and that's why we fight for our products and that's why our products are growing again, you can repeat all these numbers over and over again, how do you get 68% growth in newest com with a 57 rule of 40, how do you do.

Speaker Change: Nearly double year customer account Youre deal count your customer count all these numbers.

Mariana Perez Mora: You can do it by having full alignment and the best products in the world, or you can do it by having weird things that barely work and trying to manipulate clients into buying your product until they realize it barely works because it's thin and of no value. And that's just not how we roll. It has been really, and by the way, on the last note, how do we have the best investors in the world? especially individual investors? Why do they stick with us when times are hard?

Speaker Change: You do it by having full alignment and the best products in the world.

Speaker Change: Or you can do it by having weird things that barely work in trying to manipulate clients into buying your product until they realize it barely works because its thin and have no value and that's just not how we roll it pound here. So it has been really and by the way on the last note how do we have the best investors in the world, especially.

Speaker Change: Individual investors why do why did why do they stick with US when times were hard why do why why do they fight for pounds here why do they every time someone writes something a named Powershares gone Marga <unk> doesn't have a product pound share doesn't have margins I'm sure. This quarter talent tiers growth is decelerating. Despite the fact, it's obviously mean, we're obviously crushing it in <unk>.

Mariana Perez Mora: Why do they fight for Palantir? Why do they, every time someone writes something in the name, Palantir's gone MAGA. Palantir doesn't have a product. It doesn't have margins.

Mariana Perez Mora: I'm sure this quarter Palantir's growth is decelerating despite the fact it's obviously, we're obviously crushing it in America. Palantir is too crazy. It will never be profitable.

Speaker Change: Erica.

Speaker Change: Talents here <unk> too crazy powertrain will never be profitable pound share can't IPO Pelletier Pelletier is too weird and odd to produce something that will be disruptive.

Mariana Perez Mora: Palantir can't IPO. Palantir is too weird and odd to produce something that will be disruptive. We knew it was bullshit, the people out on the front fighting against that BS, fighting against lazy, inept, discriminatory in the sense that it is ignorance. People maligning Palantir do it in great part because they believe that we are fighting for them, and we are. We're fighting for ourselves, and that's why we have such a great company. So, thank you. Thank you. That concludes. Are you ready to make your sales team unstoppable? Put Pantodoc to work.

Speaker Change: We knew was bullshit the people who are on the front fighting against that PFS fighting against lazy inept.

Speaker Change: Discriminatory in a sense that there's ignorance.

Speaker Change: People Maligning pallets, you do it in great part because they believe that we are fighting for them and we are in.

Speaker Change: And we're fighting for ourselves and that's why we have such a great company.

Speaker Change: So thank you.

Speaker Change: Thank you that concludes the Q&A for today's call.

Speaker Change: Ready to make your sales team unstoppable panted ought to work create and share stand.

Mariana Perez Mora: Create and share standout proposals in minutes. Send error-free quotes and close 36% more deals each month. Outshine, outsmart, and outbid the competition. Discover the Pantodoc Advantage. While both NetSuite and QuickBooks offer software designed to help businesses manage their accounting processes, QuickBooks primarily focuses on accounting software for small businesses.

Speaker Change: Proposals in minutes send error free quotes and closed 36% more deals with each month outshine outsmart and outbid the competition discover the payment Ark advantage.

Speaker Change: Okay.

Speaker Change: While both net suite and Quickbooks offer software designed to help businesses manage their accounting processes Quickbooks, primarily focuses on accounting software for small businesses supporting additional complexities requires layering additional systems and apps.

Mariana Perez Mora: Supporting additional complexities requires layering additional systems and applications. NetSuite, on the other hand, offers an entire suite of business management applications, extending beyond accounting and finance, and integrating customer management, e-commerce, HR, project management, inventory management, supply chain management, and more, without any manual processing. Everything will be Everything will be Everything will be In life's way Oh, but I'll never see I'll never see I'll do it all for me I'll do it all for me And I'll never see In life's way Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Oh Everything will be Everything will be Everything will be, Please silence all cell phones and electronic devices.

Speaker Change: A patient not to be on the other hand offers an entire suite of business management applications, extending beyond the accounting and finance and integrating customer management E. Commerce, HR project management inventory management supply chain management, and Martin without any manual processes.

Speaker Change: Okay.

Speaker Change: Oh.

Speaker Change: Sure.

Speaker Change: Yes.

Speaker Change: A year ago.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: There was.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Yeah.

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Yes.

Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: Yeah.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Earlier.

Speaker Change: Yes.

Speaker Change: <unk>.

Speaker Change: Yes.

Speaker Change: Okay.

Right.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: <unk>.

Speaker Change: Right.

Speaker Change: <unk>.

Speaker Change: Please silence all cell phones and electronic devices. The program is about to begin.

Speaker Change: The program is about to begin. Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech. Please welcome from Palantir, Sasha Spivak. Hi, everybody. We're so thrilled to welcome you back to AIPcon. And on a week of particularly big milestones for us, not only for all of the commercial customers that we have gathered in this room today, but also for the US government business, as we celebrate the Titan Award and becoming the first software prime on a hardware system. Thank you. Thank you. We've had nearly 850 bootcamps completed since the launch of the program, more than 40 with hospitals and health systems.

Speaker Change: Yeah.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Please welcome from Palin tier Sasha Spivak.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Hi, everybody. We're so thrilled to welcome you back to AIP Con.

Speaker Change: And on a week of particularly big milestone for us not only for all of the commercial customers that we have gathered in the room here today, but also for the U S government business as we celebrate the tightened award and becoming the first software prime on a hardware system.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: We've had nearly 850 boot camps completed since the launch of the program more than 40 with hospitals and health systems last week alone every single customer that is with US today has completed a bootcamp before coming.

Unknown Attendee: Last week alone, every single customer that is with us today has completed a bootcamp before coming. And this afternoon, for the first time, we have customers leading their own bootcamps with each other. We've had a pretty relentless focus on AI in production the last few months in particular. Not AI on a whiteboard, not AI in a slide deck, but AI in the theater and on the factory floor. And today, we're so excited to have more speakers, individuals, customers, examples, use cases, ideas, and learnings than we've ever had. Thank you so much for being here. Thank you for tuning in. We hope you enjoy the show.

Speaker Change: And this afternoon for the first time, we have customer is leaving their own bootcamp with each other.

Speaker Change: We've had a pretty relentless focus on AI and production the last few months in particular.

Speaker Change: <unk> AI on a whiteboard not AI in our slide deck, but AI in theater and on the factory floor and today. We are so excited to have more speakers individuals' customers example, use cases ideas learnings and we've ever had.

Speaker Change: Thank you so much for being here. Thank you for tuning in we hope you enjoy the day and with that I'd like to introduce Alex.

Unknown Attendee: And with that, I'd like to introduce Alex. Welcome back to Great Day Washington. This segment is sponsored by Legacy Tax. IRS, excuse me, Fresh Start Initiative. What is that? Is that some good news, finally, Corey? Under this program, those that qualify can see their balances reduced by thousands or even tens of thousands.

Speaker Change: Okay.

Alexander C. Karp: Welcome back to Great Day, Washington. This segment is sponsored by legacy tax higher Ethics gives me fresh start initiative, what does that is that some good new find Macquarie under this program those that quantify can see their balances reduced by one thousands or even tens of thousands and some people are literally seeing 80%, 90% reduction of what they owe through the fresh start program, but it's only available.

Alexander C. Karp: Global for a limited time and only for those that qualify my team and I here at legacy tax has helped many many people throughout Maryland, DC and Virginia.

Alexander C. Karp: Some people are literally seeing an 80 to 90% reduction in what they owe through the Fresh Start program. But it's only available for a limited time, and only for those that qualify. My team and I here at Legacy Tax have helped many, many people throughout Maryland, D.C., and Virginia get that IRS Fresh Start. So if you owe the IRS and you can't afford to pay your balance in full, the Fresh Start Initiative is an opportunity you'll want to take advantage of.

Speaker Change: Fresh start so if you own the IRS and you can't afford to pay your balance and all the fresh start initiative is an opportunity youll want to take advantage of Wow. Finally, some goodness I'm gonna ended on that corner. Thank you for all of your information today. We appreciate it and if you owe the IRS and you would like to see if you qualify for a fresh start scoring hankerson has an online qualification tool at <unk>.

Speaker Change: Wow. Finally, some good news. I'm going to end it on that, Corey.

Corey: Thank you for all of your information today. We appreciate it. And if you owe the IRS and you would like to see if you qualify for a Fresh Start, Corey Hankerson has an online qualification tool at levyking.com.

Speaker Change: Tom just answer a few simple questions and you will be on your way you can also reach out to the legacy tax by calling 187 hundred 75389546.

Corey: Just answer a few simple questions, and you will be on your way. You can also reach out to Legacy Tax by calling 1-877-538-9546. Very happy to be here. This is also an anniversary because, usually, they don't let me on stage without someone to mind me. That's the whole reason. Thank you to people who are here for the second, third time. You may notice I've escaped from my minder, and the person interviewing me used to be a high-ranking official in the U.S. government. And he doesn't tell you that.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: The fourth quarter.

Speaker Change: Okay.

Very happy to be here is also an anniversary because usually they don't let me on stage without someone to mind me, let's go.

Speaker Change: Reason, thank you for people here for the second and third time, you may notice I have escaped my minder and the person interviewing me as we used to be a high ranking official of the U S government and he doesn't tell you that treaty to reprimand me to some unknown place if I go offline, but.

Corey: And he's ready to reprimand me to some unknown place if I go offline. But in any case, it is a momentous occasion for Palantir and, I would say, for our journey. And that journey began not very far from here. I don't mean the time Peter and I were in the dorm fighting with each other, although that's where it began at Stanford.

Speaker Change: Any case.

Speaker Change: It is momentous occasion for talent here.

Speaker Change: And I would say for our journey and I would.

Speaker Change: And that journey.

Speaker Change: Began not very far from here.

Speaker Change: I don't mean, the time, Peter and me in the dorm fighting with each other that's where it began at Stanford.

Speaker Change: And it really began with a central set of ideas of course, the most important one of which was we've got to make our institutions actually work, which means you put more or less.

Corey: And it really began with a central set of ideas, of course, the most important one of which was that we've got to make our institutions actually work, which means you put less into the system than you get out. A very simple idea. Every business in this room works on that internally. You build something efficiently that's better than your competition in a way where you get a margin. And that idea was in Silicon Valley, one of the building blocks of how Silicon Valley was built; it meant you built things for the US government, they actually became valuable, and then you sold them primarily to America and the rest of the world.

Speaker Change: Less into the system, then you get out a very simple idea every business in this room works on that internally you build something.

Speaker Change: Efficiently that's better than your competition in a way where you.

Speaker Change: Get a margin.

Speaker Change: And that idea.

Speaker Change: Was in Silicon Valley one.

Speaker Change: Was the building block of of how Silicon Valley was built.

Speaker Change: It meant you built things for the U S government, they actually became valuable and then you sold them too.

Corey: And because you had done all this high-end R&D, and not just done it, but done it in a way where it actually worked with the most interesting, innovative people in the world doing it, you could take those products to the world. And that actually is what built a lot of what made things work in this country really well and brought our country together in so many ways and has led to a dominance in, it's like this weird thing where 84% of the top 50 companies in the world are American, and almost all of them were started here.

Speaker Change: <unk> America, and the rest of the World and because you had done all of this high end.

Speaker Change: R&D and not just done it but done it in a way where it actually work with the most interesting innovative people in the world doing it you could take those products to the world and that that actually is what built a lot of what made things work in this country really well and broader country together and so many.

Speaker Change: Ways and has led to a dominance.

In.

Speaker Change: This weird thing, where 84% of the top 50 companies in the World, Our American and almost all of them that were started here.

Speaker Change: And.

Corey: And, And then ways of doing things, ways of working together, ways of innovating that were unique. And then, of course, we lost that and had a wave where basically people built technologies where the person they're selling it to was de facto the host, and the people selling it were de facto parasitic. And that's what the consumer Internet basically is. You enjoy yourself, you enjoy yourself, they make all the money, and you get slower and dumber, and you enjoy it, and this company and the products we built.

Speaker Change: And then ways of doing things ways of working together ways of innovating that were unique and then of course, we lost that and had a wave where basically people built technologies, where the person they're selling at too was the facto the host and the people selling it were to facto parasitic.

Speaker Change: And that's what consumer Internet basically as you enjoy yourself you enjoy yourself they make all their money and you get slower and dumber.

Speaker Change: And you enjoy it.

Speaker Change: It sounds like I'm getting slower in getting dumber and everyday I enjoy it more somehow.

Speaker Change: The slowness the slowness the slowness exercise for my mind don't need that.

Speaker Change: And.

Speaker Change: And this company and the products we build.

Corey: Leaving aside the kind of pro-West thing that I'm very proud of, it was built on the simple idea that our U.S. government clients, primarily, although we have allies from almost every country in the world using our product, would get more value from the product than what they were paying for. It would not only make their institution work, but it would work better.

Speaker Change: Leaving aside the kind of pro West thing that I'm very proud of was built on a simple idea that our U S government clients, primarily although we are allies from almost every country in the world using our product would get more value from the product than what they were paying for it would not only make their institution work, but it would.

Corey: So the first product we built was built to increase civil liberties and decrease terrorism, at a cost that is very small for what we built. Now, if you fast forward to now, when we launched, first we launched this product foundry. And then we built some things in the U.S. government and commercially, Theontology, which is now becoming famous, other parts of our product. The whole reason these products were built years before anyone thought of large language models was that we built a culture around, well, what should an enterprise do? What ought an enterprise want?

Speaker Change: Work better so the first product rebuilt was built increased civil liberties and decrease terrorism.

Speaker Change: At a cost that is very small for what we built.

Speaker Change: Now if you fast forward to now.

When we launched first relaunched this product foundry.

Speaker Change: And then we built some things in the U S government and commercially the ontology, which is now becoming famous other parts of our product. The whole reason. These products were built years before anyone thought of large language models.

Speaker Change: Was that we built a culture around well, what Shannon enterprise, what ought to an enterprise want if you took the enterprise and reduced it to what it ought to want to do with its software with it ought to want to do with its production.

Corey: If you took the enterprise and reduced it to what it ought to want to do with its software, what it ought to want to do with its production, what it ought to want to do on the battlefield, meaning, as was mentioned, the battlefield doesn't even work unless it's software imagined and generated.

Speaker Change: Art I wanted to do on the battlefield, meaning.

Speaker Change: It was mentioned at the battlefield doesn't even work unless it's software imagined and generated and then in the commercial context, how would you deal with a very harsh competitive environment.

Corey: And then in a commercial context, how would you deal with a very harsh competitive environment where you have varying quality of assets, supply chains being disrupted, you have to make people work in manufacturing as if they're from a different country. In this country, you have to reduce your margin, you have to reimagine your supply chains. Get rid of the PowerPoint.

Speaker Change: Where you have varying quality of assets supply chain is being disrupted you have to make people.

Speaker Change: Work in manufacturing as if they're from a different country in this country.

You have to reduce your margin you have to re imagine their supply chains and.

Speaker Change: What would the attributes of a software platform look like and that's how we came up with foundry and the ontology and all these things it almost looks like magic that they now exists you have this incredible Revolution and then the revolution is really confusing.

Speaker Change: Confusing because.

Speaker Change: Some large language models create value large language modestly certain things well they don't do other things well you have Powerpoint production, which is a plate on our society.

Speaker Change: Get rid of the Powerpoint that we apparently are do things, mostly for good reasons, but I not anywhere else a pound here also do things to humiliate people selling powerpoints.

Corey: We at Palantir do things mostly for good reasons. But I, not anyone else at Palantir, also do things to humiliate people selling PowerPoints. No one at Palantir would do this, but I do it, and I love it. The whole reason for the boot camp is to, one, show you something in action, something that is yours and that you own that creates enormous value for you, more value for you than what you're paying us, and shows you how you can reimagine your enterprise.

Speaker Change: No one apology reduced but I do it and I love It the whole reason for the boot camp is two one show you something in action is something that is yours in that you own that creates enormous value for you more value for you than what youre paying us <unk>.

Speaker Change: And shows you how you can re imagine your enterprise by the way, it's an enormous advantage in this country because people are maximally willing to change learn be meritocratic compared to any other country in the world and so it gives them a disadvantage to America.

Corey: By the way, it's an enormous advantage in this country because people are maximally willing to change, learn, and be meritocratic compared to any other country in the world. And so it gives a disadvantage to America. It is a disadvantage to the rest of the world.

Speaker Change: Is a disadvantage to the rest of the world is a massive advantage to America industrial base in American can be re imagined enterprises can be rebuilt the rest of the world is not moving as quick for lots of reasons. We could go into it's a massive advantage for our warfighters people protecting us because they can re imagine and I have to admit it's very humiliating to people.

Corey: It's a massive advantage to America. The industrial base in America can be reimagined. Enterprises can be rebuilt. The rest of the world is not moving this quick for lots of reasons we could go into.

Corey: It's a massive advantage for our war fighters, people protecting us because they can reimagine it. And I have to admit it's very humiliating for people selling PowerPoints and steak dinners, which I love. And I think, by the way, people, you guys, are also tired of you being able to get a steak dinner yourself. You don't need the steak dinner.

Selling powerpoints and steak dinners, Jai love and I think by the way with people. You guys are also tired of you can get a state getting yourself, what do you need to stay tuned.

Corey: You can all buy it yourselves. How about your enterprise reimagined and no steak dinner? That's what Palantir is offering. I find it super gratifying that we at Palantir and you, our partners, many of whom I've met here or will meet, are playing an enormous role in a crazy and interesting revolution, much bigger. You know, sort of the other thing, the reason why people don't realize the revolution is that people assume a revolution is associated with long, off-colored hair and psychedelic drugs.

Speaker Change: All combined yourself, how about your enterprise re imagined and no state Kinner, that's what powershares offerings.

Speaker Change: Sure.

Speaker Change: I find it Super gratifying that we at <unk> and you are partners.

Speaker Change: Many of whom I've I've met here or will meet are playing an enormous role in a crazy and interesting revolution much bigger it's really the other thing. The reason why people don't realize the revolution is people assume a revolution is associated with long off colored hair and psychedelic drugs.

Speaker Change: People hear happen.

But it's not like actually this revolution is being driven by people in this room and it is going to have an enormous impact on America on the rest of the world and at the other thing about the suite. That's really caused mostly since a lot of these things we will talk about werent really true you could wait, but what really is happening.

Speaker Change: Here's the early movers are going to move much quicker and much faster and those by the way we have a whole bunch of people in London, who are here. So this is like America, England. All these places galvanizing, but the early movers are going to have a enormous advantage over the people that wait precisely because the incremental.

Corey: And those, by the way, we have a whole bunch of people in London who are here. So this is like America, England, all these places galvanizing. But the early movers are going to have an enormous advantage over the people that wait precisely because the incremental value is not incremental. And it's like every day you wait is a day where you can't catch up.

Speaker Change: <unk> is not incremental and it's like.

Speaker Change: Everyday you weight is a day, where you can't catch up that's the advantage of Silicon Valley over the rest of America over the rest of the world. That's the advantage of pound here given that we built these things and quite frankly, that's the advantage of everyone's sitting in this room that you can actually begin to re imagine rebuild your enterprise in hours.

Corey: That's the advantage of Silicon Valley and America over the rest of the world. That's the advantage of Palantir, given that we've built these things. And quite frankly, that's the advantage of everyone sitting in this room, that you can actually begin to reimagine, rebuild your enterprise in hours, while somebody else around the world is hearing from somebody telling him this stuff isn't real. And with that, I'd like to welcome you

Speaker Change: While somebody else around the world is hearing from somebody is telling them this stuff isn't real and with that I'd like to welcome you. Its tremendously exciting for me, it's like a real high point.

Corey: It's tremendously exciting. For me, it's like a real high point to be back in Palo Alto with such a tremendous group of people and my Palantir colleagues. We have a really good vibe on our end, you know, and most people at Palantir didn't get to do a lot of winning in high school. So it's kind of, kind of nice, kind of nice, it's a new vibe, it's a new vibe, it's a new vibe.

Speaker Change: To be back.

Speaker Change: In Palo Alto with such a tremendous group of people.

And my Powersharing colleagues, we have a really good vibe on our end.

You know it's in most people a pound here didn't get to do a lot of winning in high school. So it's kind of.

Speaker Change: Kind of a nice kind of nice.

Speaker Change: It's a new vibe isn't nearby.

Speaker Change: It's a new vibe.

Speaker Change: Yeah. So, in any case, with that, very, very welcome, very happy you're here, and yeah. Glad to be partnered with you. Whatever you do, do it for less at Harbor Freight.

Speaker Change: So any case with that where are you where you are welcome very happier here and yeah glad to be partnered with you.

Whatever you do do it for less at.

At Harbor freight.

Speaker Change: Please welcome the winner of the Most Impactful AI Implementation from Lenar, Scott Spradley. Good morning. You guys should stand up and look around at how crowded this is because I think a couple years ago I was here and it was smaller, but this is humongous, and I love the quaintness just to give you the feedback, Sasha, that keeping this tight like this is the way it should be done. So thank you for that. Anyway, I'm going to tell you guys a quick story about Lenar.

[music].

Speaker Change: <unk> block on the winner of the less impactful AI implementation, rather linear Scott Spradley.

Speaker Change: Okay.

Speaker Change: Thanks.

Speaker Change: Okay.

Speaker Change: Good morning, you guys should stand up and look around at how crowded. This is because I think a couple of years ago was here.

Speaker Change: And it was smaller but this is humongous and I love. The quaintness just to give you the feedback Sasha that keeping this tight like this is the way it should be done so thank you for that anyway.

Speaker Change: I'm going to tell you guys a quick story.

Scott Spradley: Lenar, if you don't know, we're the biggest home builder in the country, and our business model is generally pretty simple to understand. So we buy land, prepare the areas that we build for construction, and then build the homes and then warranty the homes. And what we're going to talk about here today is a really complex part of that process, which is what we call horizontal construction, and that's where we're preparing the land for construction. I'm going to show you some emails here, and you don't need to read these emails, trust me. Emails are horrible.

Speaker Change: When are when or if you don't know where the biggest homebuilder in the country.

Speaker Change: And our business model is generally.

Speaker Change: Pretty simple to understand so by land.

Speaker Change: Prepare the areas that we build for construction.

Speaker Change: And then build the homes and then warranty to homes and what we're going to talk about here today is a really complex part of that process, which is what we call horizontal construction and thats where were preparing the land for construction. So.

Speaker Change: I'm going to show you some emails here and you don't need to read these e-mails Trust me emails are horrible.

Scott Spradley: But in any case, I'm just going to kind of show you a couple things. So basically, we had a guy that worked for the COO that was saying, hey, I want to buy something called Quickbase. And it's basically an Access product. Everybody in here probably groans when they hear that word because Access is not something that we want to use to feed any kind of system at all.

But in any case.

Speaker Change: I'm just going to kind of show you a couple of things. So basically we had a guy that worked for the COO that was saying Hey, I want to buy something called quick base and is basically an access product.

Speaker Change: Everybody in here, probably groans when they hear that word because access is not something that we want to use a feed any kind of system at all.

Scott Spradley: And so basically... One of the big problems that we face when we're doing that is you'll get like seven different bids, and there will be thousands of items in these bids where you're trying to find out what's the best price. So we have 17,000 different communities across the country. That's a lot, right?

Speaker Change: And so basically.

Speaker Change: This is one of the big problems that we face when we're doing that is you'll get like seven different bids and there will be thousands of items in these bids.

Speaker Change: Where you are trying to find out what's the best pricing. So we have 17000 different communities across the country.

Speaker Change: That's a lot right last year, we built we plan to build 62000 homes. We built 73000 homes. So that was kind of a big achievement, but theres a lot of complexity in preparing these homes for what we call that horizontal construction piece and so.

Scott Spradley: Last year we built, we planned to build 62,000 homes, but we built 73,000 homes. So that was kind of a big achievement. But there's a lot of complexity in preparing these homes for what we call that horizontal construction piece. And so. The COO says to me, hey, I don't know if we should buy this software or not. Can you talk to these guys?

CEO: The C O says to me I don't know if we should buy the software or not can you talk to these guys. So we talked to them and we engaged and then.

Scott Spradley: So we talked to him, and we engaged. And then the COO sent back the thing that we all dread: need speed.

Hello since back the thing that we all dread need speed need this to happen fast and that's something that is always annoying to people, who deliver technology because technology doesn't usually happen that fast but in any case so.

Scott Spradley: I need this to happen fast, and that's something that is always annoying to people who deliver technology because technology doesn't usually happen that fast. But in any case, so. Lee Slazak is here.

Scott Spradley: We sat down, and we had a conversation with this guy that wanted to buy this piece of software. And we kind of listened to him, listened to the problem that he had. And basically, one of the problems that he has is that he kind of reiterated this thing about, You know, there may sometimes be sometimes seven different bidders that provide all of these thousands of bids per item. So when you think about, like, a new home,

CEO: At least laser accurate here, we sat down we had a conversation with this guy that wanted to buy this piece of software and we kind of listen to them listen to the problem that he had to solve and basically one of the problems that he has he kind of reiterated this thing about.

CEO: There may be sometimes seven different bidders.

CEO: That provide all of these thousands of bids per item. So when you think about like a new home sometimes.

Scott Spradley: Sometimes there's manhole covers, there's 2x4s, there's plumbing, there's electricity, there's all the stuff that goes in, and it literally is thousands of lines. And these guys were doing this on spreadsheets, which, you know, by the way, I have shirts that say, "Did you not know Excel is dead?" But in any case, we thought, my God, and so as we kind of walked through this whole process with him, we thought, jeez. This is really the ultimate case for foundry. And we just kind of looked at each other, and we were like,

CEO: Sometimes theres manhole covers theres two by fours theirs.

CEO: Theres plumbing, there's electricity, there's all of the stuff that goes in and literally thousands of lines and these guys were doing this on spreadsheets.

CEO: Which by the way I have shirts that say did you not know <unk> is dead.

CEO: But in any case.

I thought my God, and so as we kind of walked through this whole process with him.

CEO: We thought Gee.

CEO: This is really the ultimate case for foundry.

CEO: And.

CEO: We just kind of looked at each other and we're like.

Scott Spradley: Why not Foundry? This is a perfect use case, and this is a great way to bring in Palantir.

CEO: Why not foundry. This is perfect use case and this is a great way to bring in talented <unk>.

Scott Spradley: I spoke here a few years ago, and we did a big transformation when I was the CTO at Tyson Foods. We took $700 million of cost out using Foundry, pretty epic, so I had a lot of confidence in that.

CEO: <unk> I had.

CEO: I spoke here a few years back and we did have a big transformation, Rob was the CTO at Tyson foods.

CEO: We took $700 million of cost out using foundry, which is pretty pretty epic so have a lot of confidence in that.

CEO: And you guys have all probably learned about foundry.

Scott Spradley: And you guys have probably all learned about Foundry. We started to work, we brought Palantir in, and in 51 days, we felt like we were ready to now show these guys.

CEO: We started to work we bought <unk>.

CEO: We had 51 days felt like we're ready to now show these guys.

Scott Spradley: What Foundry can do, what Palantir can do. And so we sat down, we asked for 40 minutes to give this demo. They said, well, you're going to get 20. Classic, right?

CEO: What foundry can do what palin peer can do.

CEO: And so we sat down we ask for 40 minutes to give this demo.

CEO: Well youre going to get 20 classic right. That's what we all live with.

Scott Spradley: That's what we all live for. [inaudible] So, we started giving the demo. Actually, Milo, who's here, was the one that was doing the demo. It was totally quiet for 14 minutes.

CEO: So we started given the demo.

CEO: Actually Milo Who's here was one that was doing the demo.

CEO: <unk>.

CEO: It was totally quiet for 14 minutes people were just like kind of an all.

Scott Spradley: People were just kind of in awe. And then, while this slide says there was an eruption, the real word is all total hell broke loose because people just completely couldn't believe what they had seen because you're basically looking through all that data. And it's showing literally that across 17,000 different communities, when you're preparing land for construction, you're saving anywhere between $2.5 million to up to $8 million per each of those communities.

CEO: And then while this slide says there was an eruption the real word is all total hell broke loose.

CEO: Because people were.

CEO: Just completely Couldnt believe what they had seen because youre basically looking through all of that data.

CEO: And it's showing literally that.

CEO: Across 17000 different communities when you're preparing land for construction.

CEO: That you are saving anywhere between two and a half million to up to $8 million per each of those communities.

Scott Spradley: It's a big deal when you think about the scale of 73,000 homes, and basically, what made us feel great was that the COO said, "It's amazing that after only 51 days, you all have changed the complete idea of technology delivery here at Lenar because we've never seen anything like it." We've got an LLM in there, we have AI running in there, all this is running, and it's basically using the LLM to create an email to just go back to each of those bidders and say, we think you can do better on this many price items because somebody else had a lower price. And that's just fantastic. So, that for us was huge, and I think Lenar has a different level of confidence. And so, my final thing that I would say is just get started. If you haven't considered it,

CEO: Big deal when you think about the scale of 73000 homes.

Basically what made us feel great is that that Cielo said, it's amazing that after only 51 days you all have changed the complete.

CEO: Idea of technology delivery here Alan R.

CEO: Because we've never seen anything like this and we've got a L. L. M and there are we have AI running in there all of this is running and is basically using <unk> to create an email to just go back to each of those bidders and say we think you can do better on this many pricing items, because somebody else had a lower price and thats.

CEO: Just fantastic so.

CEO: That for US was huge and I think <unk> has a different level of confidence and so my final thing that I would say is just get started.

CEO: If you haven't considered if you're a customer in here and you're thinking about using Lynn or feel free to give me a call because I can tell you what the velocity that youre going to get the intellect of the people that are there the amount of learning that theyre going to do as far as the business case goes.

Scott Spradley: If you're a customer here and you're thinking about using Lenar, feel free to give me a call because I can tell you what the velocity that you're going to get, the intellect of the people that are there, the amount of learning that they're going to do as far as the business case goes is like you've never seen before. So, that's my story. Thank you guys for being here and enjoying Foundry, or excuse me, AIPCON. Please welcome the winner of the best end-to-end AI integration from Lowe's, Elena Wielden. Lucidchart makes intelligent diagramming easy and helps your best ideas become real. Start from scratch or get started with templates.

CEO: Like you've never seen before so that's my story. Thank you guys for being here and enjoy foundry.

Speaker Change: Excuse me.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Hello.

Speaker Change: Please welcome the winner of the best end to end AI integration from Loews Alena will then.

Scott Spradley: Shape your ideas with hundreds of shape libraries, from technical systems to flowcharts, process maps, and even organizational charts. Lucidchart integrates with your favorite software applications, so it works where you work. Share documents with teammates. Commenting and tagging make it easy to align with your team. Understand the past so you can build your future. Ditch complexity; embrace clarity. Start diagramming today with Lucidchart. All right, how many of you have had a home improvement project or bought something? All right, you better have bought it at Lowe's, and if not, you better download the Lowe's app right now or maybe a little bit later.

Speaker Change: [music].

Speaker Change: Perfect.

Speaker Change: Lucid chart makes intelligent diagramming easy and helps your best ideas become REO.

Speaker Change: Start from scratch or get started with temporary.

Speaker Change: Shape your ideas with hundreds of shape libraries from technical systems to flow charts of process maps and even Org chart.

Speaker Change: Lucid chart integrates with your favorite software applications. So it works where you work.

Speaker Change: Okay.

Speaker Change: Share documents with teammates, commenting and tagging make it easy to align with your team.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: I understand the past so you can build your future.

Speaker Change: Okay.

Speaker Change: Ditch complexity.

Speaker Change: <unk> clarity.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Start diagramming today with Lucid chart.

Okay.

Speaker Change: Okay.

Speaker Change: However, I just want to say thank you so much to the Palantir crew for inviting us out here and sharing with everybody what we've been able to do in record time. So I will echo what Scott said about record speed to market. So let me tell you a little bit of a story about our home services business. As a customer, you can go into a Lowe's store or shop us on Lowes.com, and you can pick out exactly what you want to do for the project you have in mind, whether or not it's a kitchen, a bath, or a flooring project.

Speaker Change: Alright, how many of you have had a home improvement project or bought something alright, you better have bought it at Lowe's and if not you better download the Lowe's App right now.

Speaker Change: Maybe a little bit later, however, I just wanted to say thank you. So much said parents here crew for inviting us out here and sharing with everybody well, we've been able to do in record time, So I will echo what Scott said about record speed to market.

Speaker Change: It could actually be a roofing project, a fencing project, you name it. Once you've checked out, essentially, it moves to a team of about 1,700 people, and it's a call center. They monitor all of the activities to take that to actual completion for our customers. We've had some, I'm going to call it, dated ERP solution that the team has been working with. And with dated you can imagine this, a lot of spreadsheets.

Speaker Change: So let me tell you a little bit of a story about our home services like business as a customer you can go into a lowes store or shop us on <unk> Dot com and you can pick out.

Speaker Change: Exactly what you want to do for the project you have in mind, whether or not at the kitchen and Bath flooring project. It could actually be a roofing project Assenting project you name it.

Speaker Change: You checked out essentially lose to a team that is about 1700 people and it's a call center. They monitor all of the activity is to take that to actual completion for our customers.

Speaker Change: We've had some I'm going to call. It dated ERP solution that the team has been working with and with David I'm sure you can imagine ness.

Speaker Change: A lot of spreadsheets.

Speaker Change: A lot of access databases sitting on somebody's machine. And you can imagine that with a remote crew, that makes it really difficult to actually audit the business itself to make sure that projects are getting completed on time and customers are getting exactly what they need. So last year, we actually began working with Palantir on another use case for the supply chain. But I'm not here to talk to you about that one.

Speaker Change: A lot of access databases, some sitting on somebody's machine.

Speaker Change: And you can imagine with a remote krill that makes it really difficult to actually audit the business itself to make sure that projects are getting completed on time and customers are getting exactly what they needed.

Speaker Change: So last year, we actually began working with Palin tier on another use case for it supply chain and I'm not here to talk to you about that one as we relay learning really quickly from our supply chain used case, and we understood. What we wanted to do with our home services business.

Speaker Change: As we were learning really quickly from our supply chain use case, and we understood what we wanted to do with our home services business, we actually said, wait a second, we have another use case for you. And so we actually rallied the troops, and. Let me just walk you through what happened next.

Speaker Change: We actually said wait a second we have another use case for you and so we actually rally the troops.

Speaker Change: And let.

Speaker Change: Let me just walk you through what happened next.

Speaker Change: So there was a crew from Palantir that was from Oslo and London that came into Charlotte, North Carolina, began to work with us to understand our problems, and then we actually used Foundry to build in the data, stitch together ontologies, and we created a POC in record time. And because of what we saw in the POC, we actually said we were able to reduce overdue activities. Overdue activities? Just think of the word. Overdue, not good.

Speaker Change: So there is a crew from talent here that were from Oslo in London that came into Charlotte North Carolina began to work with us and understanding our problems.

Speaker Change: And then we actually used foundry to build and later data stitched together ontologies and we created a POC and record time.

Speaker Change: And because of what we saw in the POC. We actually said, we were able to reduce the overdue activities overdue activities just think of the word.

Speaker Change: And essentially, a 75% reduction in those activities, which means we were improving our customer service. On top of that, people were becoming more productive. And on top of that, managers had more insight into what was actually happening with the business itself. And in less than four months, we created something from POC all the way to production. And I'm hoping Alex is in London, and we've reached 1,000 people a day.

Speaker Change: <unk> not good and essentially 75% reduction in those activities, which means we were improving our customer service on top of that people really becoming more productive and on top of that managers had more like insight into what was actually happening with like the business itself.

Speaker Change: And lastly in four months, we created something from POC, all the way to production.

Speaker Change: And.

Speaker Change: I'm, hoping Alex is in London, and we've reached a thousand people a day, but essentially on average we have a thousand people who are working.

Speaker Change: But essentially, on average, we have 1,000 people who are working. Simultaneously, you can see where we've been, but we launch. And I will tell you, the call center agents, or what we like to call our services, like organization, they're so excited about this, like new technology. And Alex said, like, you know, making us slower and dumber, like Palantir has made us faster and smarter in general. And one of the things that, you know, you always pride yourself on is how fast can I do something to make somebody happy.

Speaker Change:

Speaker Change: Simultaneously you can see like where we've been but we launched and I will tell you. The call Center agents are what we like to say is our.

Speaker Change: Services like organization Berlin. So excited about this like new talent technology, and Alex I'd like making a slower in tomo like talent here has made us faster and smarter and genre and one of the things that.

Speaker Change: And the business itself couldn't believe how fast we were able to do that. And so, with that, there's a lot more work ahead of us for sure. And, uh, as you can imagine, now that the business is all right, we have the ability to have a bird's eye view, prioritize like workflows, and essentially understand the productivity and health of our like network. What else can we do?

Speaker Change: You've always prided yourself on is how fast can I do something to make somebody happy and the business itself <unk>, how fast we were able to do that and.

Speaker Change: So with that there's a lot more work ahead of us for sure and as you can imagine now that the business has all right. We have the ability to have bird's eye view prioritize workflows.

Speaker Change: Essentially understand the productivity and health of our like network what else can we do this so we're thinking how do we like to take AIP to the next level so that when a.

Speaker Change: So we're thinking, how do we take AIP to the next level so that when a call center agent needs to contact a provider who hasn't, you know, shown up for a job or things are running behind, we're actually sending that person a little nudge, and it's not a person calling at the end of the day. So we are just now getting started on our journey, but I will tell you this, without Palantir, we definitely would not have been able to solve this as quickly as possible. And with that, Alex, if you're watching in London, thank you so much for being part of the team.

Speaker Change: I'll call center agent needs to contact a provider who has shown out for job where things are running behind we're actually finding that person a little niche and it's not a person calling at the end of the day. So we are just now getting started on our journey, but I will tell you that.

Speaker Change: Without talent tier, we definitely would not have been able to solve this as quickly as possible and with that.

Speaker Change: Alex if youre watching in London. Thank you so much for being part of the team I Echo everything that Scott said is that the team rallied and literally showed up wearing the red box for us. Thank you.

Speaker Change: I echo everything that Scott said, and the team rallied and literally showed up wearing the red vest for us. Thank you. Please welcome the winner of the most ambitious AI roadmap from Archer Aviation, Adam Wormuth. I was 17 years old when I took over my family's small farm.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Please welcome the winner of the most ambitious AI roadmap from Archer Aviation Adam warm F.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: I was 17 years old when I took over my family small farm.

Adam Wormuth: Today, we farm over 15,000 acres and grow the majority of pumpkins sold in America. And that's why I use NetSuite by Oracle. It gives me a full picture of my business, from accounting to inventory, supply chain management, HR, and so much more. I get all of this information, all in one place, whenever I need it. Hardworking Americans use NetSuite, do you? Visit netsuite.com slash farms for a product tour. All right, all right. Thank you Palantir for having me here today. As an ex-Anderillian, congrats on the Titan win as well.

Speaker Change: Today, we farm over 15000 acres and grow the majority of Pumpkin sold in America, and that's why I use naturally <unk>.

Speaker Change: It gives me a full picture of my business from the county to inventory supply chain management, HR and so much more.

Speaker Change: All of this information all in one place whenever I need it.

Speaker Change: Hardworking Americans using ethylene.

Speaker Change: Visit net suite Dot com slash farms for a product tour.

Speaker Change: Okay.

Speaker Change: Really exciting news. So I'm here today to tell you about Archer Aviation, our midnight aircraft, and how we're working together with Palantir AIPE to transform urban mobility. How many people were stuck sitting in traffic on the 101 this morning trying to get here?

Alright. Thank.

Speaker Change: Thank you pound here for having me here today as an actor Andrew William Congrats on the tightened win as well really exciting news, though.

Speaker Change: I'm here today to tell you about Archer aviation our midnight aircraft.

Speaker Change: And how we're working together with pound here AIP to transformer mobility, how many people were stuck sitting in traffic on the one on one this morning trying to get here a lot of folks.

Speaker Change: Yeah, a lot of folks. And I'm sure you're sitting there wondering, "yo, why are we still doing this?" It's the 21st century; there has to be a better way. And Peter Thiel famously said, "We wanted flying cars, and we got 104 characters, 140 characters." But I'm here to tell you today, the future is coming, and Archer is going to be launching next year to transform it. So Archer's mission is really about how we make our cities more efficient, better places to live and work. And so half of the world's population today lives in cities; it's going to be two-thirds by 2050.

Speaker Change: And I'm sure you're sitting there wondering.

Speaker Change: Why are we still doing this 20 <unk> century, there has to be a better way and Peter T. O famously said we wanted.

Speaker Change: Buying cars and we got 104 characters 140 characters, but I'm here to tell you today.

Coming in Orange is going to be launching next year to transform itself.

Speaker Change: So our true mission is really about how do we make our cities more efficient better places to live and work.

Speaker Change: So half of the world's population today lives in cities is going to be two thirds by 2050 and as our populations have grown our economic activity has grown our cities has been able to keep up with that growth by scaling into the third dimension and building more buildings skyscrapers, but our transportation is really been stuck in two dimensions in a gridlock.

Speaker Change: And as our populations have grown, our economic activity has grown, our city has been able to keep up with that growth by scaling into the third dimension and building more buildings, skyscrapers, but our transportation has really been stuck in two dimensions in a gridlock. And so, as our population grows, there's really not any place to grow and nowhere to go except for up. And so we look at cities like San Francisco and San Jose, Chicago, New York and ask why aren't we flying around these cities today? I flew out here from New York on Tuesday, and I very nearly missed my flight, getting stuck in the gridlock trying to get into the Holland Tunnel.

Speaker Change: So as our population grows there's really not any <unk>.

Speaker Change: Ways to grow and nowhere nowhere to go except for up and so we look at cities like San Francisco and San Jose.

Speaker Change: Cargo, New York and looked at why aren't refine around these cities today.

Speaker Change: I flew out here from New York on Tuesday, very narrowly missed my flight getting stuck in the grid lock trying to get into the Holland tunnel.

Speaker Change: And so we've got this, you know, massive city, but, you know, everyone trying to go through one pinch point. So when you look at going to the third dimension, taking advantage of all that space, you could fly there in just nine minutes. And the technology that's going to enable us to do this really at scale is the electric propulsion that drives Archer's Midnight aircraft. And so because Archer is all electric, it's zero emissions, it's 100 times quieter than a helicopter flying overhead, and it's fully redundant, which is going to make it as safe as flying in a commercial aircraft today.

Speaker Change: So we've got this massive city, but everyone trying to go through one one pinch point. So when you look at go into the third dimension taken advantage of all of that space you could fly there and just nine minutes.

Speaker Change: And the technology, that's going to enable us to do this really at scale is the electric propulsion that drives archer's midnight aircrafts and so because our tours all electric is zero emissions, it's 100 times quieter than a helicopter flying overhead and it's fully redundant where's it going to make it as safe as finding a commercial aircraft today and so those core pieces of electric technology.

Speaker Change: <unk> is what has it been able us to actually really go and operate this at scale at a cost that's affordable to everyone and actually go transformer mobility. So.

Speaker Change: And so those core pieces of electric technology are what is going to enable us to actually really go and operate this at scale, at a cost that's affordable to everyone, and actually transform our mobility. Archer was really designed from the first sketch around this mission, and so it's got four passengers and a pilot, flies at 150 miles an hour in a range of 100 miles, really designed for this urban transportation mission, moving people, you know, from downtown Manhattan to the airport, moving people from San Francisco to San Jose, and doing it in a way that communities are going to want and accept.

Speaker Change: Arthur was really designed from from our first sketch around this mission and so it's Scott for passengers in a pilot five to 150 miles an hour and a range of 100 miles really designed for this urban transportation mission moving people from downtown Manhattan to the airport moving people from San Francisco to San Jose and doing away that community, if theyre going to want and going to accept.

Speaker Change: <unk>.

Speaker Change: In order to do this, in order to achieve our vision, we're going to have to build aircraft at a scale an order of magnitude greater than has ever been seen in commercial aviation. And so we're working with our partner Stellantis. We're really excited to work with them. They've got people embedded in the team.

Speaker Change: In order to do this in order to achieve our vision, we are going to have to build aircraft.

Speaker Change: At a scale to order of magnitude greater than has ever been seen in commercial aviation and so we're working with our start partners want us and we're really excited work with them they've got folks embedded in the team, we're leveraging their expertise and their knowledge with with with.

Speaker Change: We're leveraging their expertise and their knowledge with Foundry and with AIP to build a facility that's going to be capable of building 2,000 aircraft a year in Covington, Georgia. And so we're really excited to leverage that knowledge, leverage AIP to really produce it on an unprecedented scale. And as we build these aircraft, then we need to operate them. And so we're looking at our cities around the world and saying, how are we going to deploy hundreds of aircraft into these cities at dozens of takeoff and landing sites, moving tens of thousands of passengers per day?

Speaker Change: Foundry and with AIP.

Speaker Change: To build a facility or can be capable of building 2000 aircraft a year in Covington, Georgia.

And so we're really excited to leverage that knowledge leveraging IP.

Speaker Change: Really produce at an unprecedented scale.

Speaker Change: And as we build these aircraft and we need to operate them and so we're looking at our cities around the world and saying how are we going to deploy hundreds of aircraft into these cities at dozens of takeoff and landing sites moving tens of thousands of passengers per day and so the complexity of this operation that that flows from how do we build our aircraft and deploy them and operate them and maintain.

Speaker Change: And so the complexity of this operation that flows from how do we build our aircraft and deploy them and operate them and maintain them, how do we handle our customer operations, how do we handle weather disruptions, all of that, to be able to do that at this sort of scale is going to require a digital twin of our entire enterprise, of all of our operations, creating a single source of truth that everyone's operating off of. You can imagine a supply chain disruption in our motor subassembly.

Speaker Change: Mmhmm, how do we handle our customer operations, how do we handle weather disruptions all of that to be able to do that at this sort of scale is going to require a digital twin of our entire enterprise of all of our operations, creating a single source of truth that everyone's operating off of.

Speaker Change: You can imagine a supply chain disruption in our motor sub assembly, we need to flow that information through to our maintenance team to say hey can we start to defer some maintenance on those items, we need to know our flight planning team to say, hey, can we adjust them or flight routes have lessened wear and tear on the motors.

Speaker Change: We need to flow that information through to our maintenance team to say, hey, can we start to defer some maintenance on those items? We need to know our flight planning team to say, hey, can we adjust some of our flight routes to have less wear and tear on the motors.

Speaker Change: We need to let our customer support team know, hey, we're going to have some schedule disruptions as a result of this, and they need to scale up their support. So a lot of these things that are highly coupled and really only possible when you connect your entire enterprise with a digital twin. And so we have already started working with Palantir today. Our midnight aircraft is flying just down the road in Salinas, California, going through flight tests.

Speaker Change: We needed to let our customer support team know hey, we're going to have some schedule disruptions as a result of this and maybe to scale up their support so lot of the things that are highly coupled and really only possible. When you conduct your entire enterprise with a digital twin and so we've already started working with pound here today.

Speaker Change: Our midnight aircraft is flying just down the road in Salinas, California going through flight test. We're learning we're iterating as we work towards our 2000 and twenty-five FAA certification and so as we make.

Speaker Change: We're learning, we're iterating as we work towards our 2025 FAA certification. And so as we make, as we make happy learnings and we make changes, we need our flight test team, our manufacturing team, our design release team, you know, our purchasing team, all operating from the same sheet of music in terms of the bill of materials of the aircraft, what the latest design is, what the latest results are from the test, and how we need to incorporate that in design, you know, any delays there, you know, if we have to wait a day between tests, right, if we have to wait a day for the supply chain team to know that the bill of material has changed, that's going to impact our program timelines and delay our commercialization. So we've already got the Palantir team on site.

<unk> earnings I mean make changes we need our flight test team our manufacturing team our design release team.

Speaker Change: Our purchasing team all operating from the same sheet of music in terms of the bill of materials of aircraft. What the latest design is what their latest results are from a test and in how we need to incorporate that in design.

Speaker Change: Any delay there you know if we have to wait a day between tasks right. If we have to wait a day for the supply chain team to know that the bill of materials has changed that's going to impact our program timelines and delay our commercialization. So we've already got the talented team on site.

Speaker Change: And they've been able to support our efforts. We've been able to build an ontology that takes the bill of materials of the aircraft and lets us view it in a hierarchical way that we really haven't been able to visualize it before. We're also able to see where the risk is in our bill of materials, where we need to apply more pressure on the purchasing side, and really get this all to come together in a faster timescale than ever seen for an aircraft certification program of this magnitude.

Speaker Change: And they've been able to to support our.

Speaker Change: We've been able to build an ontology that takes the bill of materials of the aircraft.

Speaker Change: This view it in a hierarchical way.

Speaker Change: No, we really haven't able to visualize it in before we're also able to then see where the risk is in our bill of materials, where we need to apply more pressure on the purchasing side and really get this all to come together in a faster time scale than ever been seen for an aircraft certification program of this.

Speaker Change: And so, you know, but we really wouldn't be working with Palantir and AIP if it was just that problem. You know, really, it's like, we've got a great point solution there, but really, it's about enabling this, you know, massive urban transportation system at scale. So thank you all for having me here today, and we can't wait for you to fly with us. Please welcome the winner of the Most Intelligent AI-Driven Supply Chain from General Mills, Dave Jackett. First, I just have to apologize to Alex. I do have three PowerPoint slides. My hope is I'm not one of those that gets humiliated up here.

Speaker Change: This magnitude and so.

Speaker Change: But we really wouldn't be working with pound here in AIP. If it was just that problem you know really it's like we got a great point solution, there, but really it's about enabling this.

Speaker Change: Massive urban transportation system at scale.

Speaker Change: So thank you all for having me here today, and we can't wait for you to fly with us.

Speaker Change: Thanks.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Please welcome the winner of the most intelligent AI driven supply chain from General Mills, Dave jacket.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: For us.

David A. Glazer: First I just have to apologize to Alex I do have three Powerpoint slides.

David A. Glazer: My Hope is I am not one of those that gets humiliated up here.

Unknown Attendee: So, a couple of years ago, I drew up a purpose statement for myself, and it was that I wanted to create beauty from chaos. And the thing that I love about my job, the thing I love about this work that we're about to talk about, is that I get a chance to do that every single day. And so I'm going to tell you a little bit about our journey at General Mills, but I wanted to start with this quote that I saw a couple of years ago. And if you're in this audience, my guess is that, on some level, this resonates with you.

So a couple of years ago I drove our purpose statement for myself and it was that I want to create beauty from chaos.

David A. Glazer: And the thing that I love about my job the thing I Love about this work that we're about to talk about is that I get a chance to do that every single day and so im going to tell you about our journey at general Mills, but I wanted to start with this quote that I saw a couple of years ago and if youre in this audience. My guess is that on some level. This resonates with you and that's that we're moving from a world where people make decisions supported by machine.

Unknown Attendee: And that's that we're moving from a world where people make decisions supported by machines to one where machines make most of the decisions guided by people. And I think at General Mills, we're on that journey, and with the help of Palantir, we're accelerating toward it. But it doesn't happen by accident.

David A. Glazer: To one where machines make most of the decisions guided by people and I think a general mills, we're on that journey with the help of pound here, we're accelerating towards it.

David A. Glazer: But it doesn't happen by accident. It takes intention it takes commitment and it takes the right partners.

Unknown Attendee: It takes intention, it takes commitment, and it takes the right partner. We started this journey back in 2019 with an investment in a connected data foundation. That's where we moved 2,000 master and operational data tables to the cloud, and connected them together. And I can't tell you how important it's been to have that single source of truth that has accelerated every single use case that's come after. That was sort of the core.

David A. Glazer: We started this journey back in 2019 with an investment in a connected data Foundation, that's where we moved 2000 master and operational data tables to the cloud connected them together and I can't tell you how important it's been to have that single source of truth.

That has accelerated every single use case that's come after it that was sort of the core then we made an investment in planning system. Many of you probably have but if you're like US you very quickly realized that our plan is just a plan and.

Unknown Attendee: Then we made an investment in a planning system. Many of you probably have, but if you're like us, you very quickly realize that a plan is just a plan, and it needs to be executed. Things change the second that they're created, and that's where intelligent execution comes in. That's what we're going to talk about, and that's where our partnership with Palantir, I think, is fundamentally changing the way that we run our supply chain.

David A. Glazer: And it needs to be executed things change the second that it's created and Thats, where intelligent execution comes in that's what we're going to talk about and that's where our partnership with found here I think is changing fundamentally the way that we run our supply chain.

Unknown Attendee: So we're building this intelligent execution layer, but we're not stopping there. The goal is eventually to connect that to our partners, so the ecosystem. These are our suppliers, these are our customers, these are our third-party logistics providers.

David A. Glazer: So we are building this intelligent execution layer, but we're not stopping there. The goal is eventually to connect that to our partners to the ecosystem. These are our suppliers. These are our customers either a third party logistics providers and so we're already starting to bring in the external data to power that side of the model.

Unknown Attendee: And so we're already starting to bring in the external data to power that side of the model. I did want to just pause briefly on this design for an intelligent execution system because I think it's really important to how this works. At General Mills, our supply chain is significantly complex, like many of you. We've got 4,000 suppliers in a network that feeds over 200 plants in North America alone.

I did want to just pause briefly on this design for an intelligent execution system, because I think it's really important to how this operates.

Speaker Change: At General Mills, our supply chain is significantly complex like many of you. We've got 4000 suppliers in our network that feed over 200 plants in North America alone, we try to service about $1 2 million customer orders every year and it's our estimate that our operational folks those we're actually touching the orders are making about $50 million.

Unknown Attendee: We try to service about 1.2 million customer orders every year, and it's our estimate that our operational folks, those who are actually touching the orders, are making about 50 million decisions every year. Those 50 million decisions are what drives the $10 billion in COGS that we have. And it's what helps us achieve or not achieve our customer service, our quality, and some of our other greenhouse gas emissions goals. What we're really trying to do with this is automate the millions of $5,000 decisions, small but mighty, that we have to make manually today every single year.

Speaker Change: Decisions every year.

Those 50 million decisions is what drives the $10 billion in Cogs that we have and it's what helps us achieve or not achieve our customer service our quality and some of our other greenhouse gas emissions goals. What we're really trying to do with this is automate the millions of 5000 dollar decisions.

Speaker Change: The small but mighty.

Speaker Change: That we have to make manually today every single year as.

Unknown Attendee: As we build that up, and this is the bottom part of it, the hope is that we're also going to create better visibility end-to-end for our leaders, that we can bring them options and ensure that their strategy is being translated to action at the ground level. That's really hard to do today.

Speaker Change: As we build that up and this is the bottom part of it. The hope is that we're also going to create better visibility end to end for our leaders.

Speaker Change: So we can bring them options and ensure that their strategy is being translated to action at the ground level.

Speaker Change: That's really hard to do today it feels like Youre trying to tell people what to do with E mails and expel spreadsheets and all of that and that's how strategy turns into action.

Unknown Attendee: It feels like you're trying to tell people what to do with emails and expel spreadsheets and all that. And that's how strategy turns into action. But there definitely is a better way.

Speaker Change: But there definitely is a better way.

Unknown Attendee: And so we're starting in our logistics space. We're starting in our raw material space. And that's what I wanted to talk about.

Speaker Change: And so we're starting in our logistics space, we're starting in our raw materials space and that's what I wanted to talk about so we started a year ago with a vision intelligent execution.

Unknown Attendee: So we started a year ago with that vision, intelligent execution, and this project called ELF for end-to-end logistics flow. And I will say the ELF branding was there before our partnership with Palantir, but it's been kind of a fun coincidence that that came together. So we started with that, and I'm happy to say that today. We've got a system that, every day in real time, is consuming constraints, capacity, and cost of our network in real time.

Speaker Change: And this project called Elf and logistics flow and I will say the Elf branding was there before our partnership with pound here, but it's been kind of a fun coincidence.

Speaker Change: That that came together.

Speaker Change: So we started with that.

Speaker Change: I am happy to say that today.

Speaker Change: We've got a system that every day in real time is consuming constraints capacity cost of our network in real time.

Unknown Attendee: Overnight, it's taking a look at over 3,000 orders. This is the movement from our plant to our warehouses, and wherever it finds disruption, wherever it finds opportunity for cost mitigation, it's making a recommendation. This is happening automatically.

Speaker Change: Overnight, it's taking a look at over 3000 orders. This is the movement from our plant to our warehouses.

Speaker Change: And wherever it finds disruption wherever it finds opportunity for cost mitigation. It is making a recommendation this is happening automatically.

Unknown Attendee: Those recommendations go to people today; eventually, we want to take them off the loop, and over 70% of those changes, those recommendations are accepted. That means that in 70% of the time. The machine is beating or exceeding what a human can do, which is not terribly surprising, but we still need that human element for now until we build the trust, but we're very close to, I think, that threshold for automation where you can start turning those decisions over directly to the machine. And it's showing real results. I mean, we're saving on average about $40,000 a day, which, if you're keeping track, that's about $14 million annually.

Those recommendations go to people today, eventually we want to take them off the loop.

Speaker Change: And over 70% of those changes those recommendations are being sited that means that 70% of the time.

Speaker Change: The machine is beating or exceeding what our human can do its not terribly surprising, but we still need that human element for now until we build the trust, but we're very close to I think that threshold for automation Rican start turning those decisions over directly to the machine.

Speaker Change: And it's showing real results I mean, we're saving on average about $40000, a day, which youre keeping track that's about $14 million annually and it's really only deploy to part of our network as we speak.

Unknown Attendee: And it's really only deployed to part of our network as we speak. We were able to build this capability in just a couple of months and deploy it because of that investment in connected data, because of the power of Foundry and Palantir from a platform standpoint, and it's resulting in decision speed, fewer touches, and eventually lower greenhouse gas emissions. I did want to say we're learning a ton on this journey, too, about change management, and the cultural shift that's needed to implement automation at scale. We're starting to build standardized adoption metrics. We're starting to upskill people and reskill them for a different way of working, and we're installing change champions in every place.

Speaker Change: We were able to build this capability and just a couple of months and deploy because of that investment in connected data because of the power of foundry and pound here from a platform standpoint, and is resulting in decision speed fewer touches and eventually lower greenhouse gas I did want to say, we're learning a ton on this journey to about change management, the cultural shift that's needed to implement automation.

Speaker Change: At scale, we're starting to build standardized adoption metrics, we're starting to Upskill people and Reskill for a different way of working and we're installing changed champions in every place I think that's critically important to it and I would say foundries impact of this is really powerful because real time, you can see the impact of the decisions, but you can also see the impact of the decisions you didn't make.

Unknown Attendee: I think that's critically important to it. And I would say Foundry's impact on this is really powerful because, in real time, you can see the impact of the decisions, but you can also see the impact of the decisions you didn't make. And that can be really powerful for teams to manage from a leadership standpoint. So I'm super excited for the work that we're doing. I'm super excited for the work that we're going to do. I'm pulling for all of you on your journeys, and I invite you to reach out if you want to go on that journey together. Thank you. One more round of applause for our first few speakers.

Speaker Change: And that can be really powerful change management leadership standpoint.

Speaker Change: So I'm Super excited for the work that we're doing I'm Super excited for that work that we're going to do I am pulling for all of you on your journeys and I offer you to reach out if you want to go on that journey together. Thank you.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: One more round of applause for our first these speakers.

Speaker Change: All right, y'all. We've got a lot more coming for you, so we're going to give you about 30 minutes to get some coffee, continue having the conversations, talk through what you learned, what you heard, what you want to hear more about. For those of you joining us on the live stream all over the U.S., all over the world, we'll see you back here in 30. Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, Please welcome the winner of Best Combination of AI and Human Expertise from Panasonic Energy North America, Tara Meisinger.

Speaker Change: Alright tell we've got a lot more coming for you. So we're going to give you about 30 minutes to get some coffee continue having the conversations talk through what you learned what you heard what you want to hear more about.

Tara Meisinger: Good morning everyone. My name is Tara Meissinger, and I want to share with you all today how Panasonic Energy is conquering tribal knowledge with the help of our friends here at Palantir. Panasonic Energy, or Pena, manufactures millions of lithium-ion batteries every day for the EV industry out of our factory in Reno, Nevada.

Speaker Change: For those of you joining us on our live stream all over the U S. All over the World, We'll see you back here in <unk>.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: [music].

Yes.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

No.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: [music] accordingly.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Alright.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: <unk>.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Okay.

Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Yes.

Yes.

Speaker Change: Yes.

Speaker Change: <unk>.

Sure.

Speaker Change: Sure.

Speaker Change: Okay.

[music].

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Sure.

Speaker Change: Yeah.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thanks.

Speaker Change: Okay.

Speaker Change: Thanks.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Sure.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Thanks.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Thank you.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Sure.

Speaker Change: Thank you.

Speaker Change: Yeah.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: <unk> count.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Thank you.

Speaker Change: Tom.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Okay.

Speaker Change: Thank you.

Speaker Change: Great.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music] okay.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Okay.

Speaker Change: Yes.

Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Okay.

[music].

Speaker Change: Okay.

Speaker Change: Thank you.

Okay.

Speaker Change: Yes.

Speaker Change: Thank you.

Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Okay.

Speaker Change: During the quarter.

Speaker Change: Correct.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Thanks.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Right.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Thanks.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Inc.

Speaker Change: Hum.

Speaker Change: [music] that.

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Uh huh.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Hum.

Speaker Change: [music].

Thank you.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Okay.

[music].

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Sure.

Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Okay.

Speaker Change: Yes.

Speaker Change: Great.

Speaker Change: Sure.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

[music].

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Okay.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: [music].

No.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Okay.

Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Thanks.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

[music].

Speaker Change: Yeah.

Speaker Change: Yes.

Okay.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Thank you.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change:

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Thank you.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: In return.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: <unk>.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Yeah.

Speaker Change: Sure.

Speaker Change: Okay.

Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Sure.

Okay.

Okay.

Speaker Change: Yes.

Speaker Change: [music] program.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Sure.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Okay.

Okay.

Speaker Change: Okay.

Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Okay.

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Sure.

Speaker Change: Sure.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Please welcome the winner of best combination of AI and human expertise from Panasonic Energy North America Tara Meisinger.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Tara Meisinger: Good morning, everyone.

Terramycin, Gerry: Everyone. My name is Terramycin, Gerry and I want to share with you all today, how Panasonic energy is conquering tribal knowledge with the help of our friends here at Palatin tier.

Terramycin, Gerry: Panasonic energy or panel manufacturers millions of lithium ion batteries every day for the EV industry out of our factory in Reno, Nevada.

Tara Meissinger: I started at Pena in 2018 in the battery engineering department before making my way over to my current role. Now I'm a senior manager for maintenance, where I oversee about 75% of our equipment repairs, and I'm responsible for approximately 350 maintenance technicians, or MTs, as we call them. Now, if I ask you to tell me what comes to mind when you think of Reno, I don't think you're going to tell me a factory or manufacturing.

Terramycin, Gerry: I started at Palomar in 2018, with the battery Engineering Department before making my way over to my current role now I am a senior manager for maintenance, where I oversee about 75% of our equipment repairs and I'm responsible for approximately 350 maintenance technicians or MTS as we call them.

Now if I ask you to tell me what comes to mind, when you think of Reno.

Speaker Change: I don't think youre going to telling me a factory or manufacturing.

Tara Meissinger: Many of our MTs come from other industries, such as entertainment, tourism, or distribution, so it's up to us to teach them a new skill set. So let's pretend we're a brand new technician in our winding process, where the anode, cathode, and separator come together to begin forming the battery.

Speaker Change: Many of our MTS come from other industries, such as entertainment tourism or distribution. So it's up to us to teach them a new skill set.

Speaker Change: So that's very moment pretend, we're a brand new technician in our widening process, where the anode cathode and separate are coming together to begin forming the battery.

Tara Meissinger: As much as I'd love to show you a real winder, my bosses are here today, so we're going to have to use our imaginations and actually the generator out back. In this demo, we're going to use mixed reality to showcase how our ontology can be leveraged to upskill a technician to solve harder problems faster. So we walk up to this winder with our trainer.

Speaker Change: As much as I'd love to show you a real Winder my bosses are here today, so we're going to have to use our imaginations and actually the generator outback.

Speaker Change: In this demo we're going to use mixed reality to showcase how our ontology can be leveraged to upskill, a technician to solve harder problems faster.

Speaker Change: So we walk up to this winder with our trainer it's.

Tara Meissinger: It's clearly not working, and we need to find the root cause. We have hundreds of machines like this on our production floor, each with tens of thousands of moving parts. Now, the benefit of being a state-of-the-art factory is that our machines output a ton of data, so much so that we ultimately don't have the ability to analyze or digest it all on our own. Excel spreadsheets just look at it and laugh. If we look at the machine data, we can see thousands of error codes on this machine, and the trainer lets you know there could be multiple solutions for every error. However, only some of the solutions are written down.

Speaker Change: It's clearly not working and we need to find the root cause we have hundreds of machines like this on their production floor, each with tens of thousands of moving parts now the benefit of being a state of the art factory is our machines output a ton of data. So much. So that we ultimately don't have the ability to analyze our digest.

Speaker Change: It all on our own excel spreadsheets, just look at it and laugh.

If we look at their machine data, we can see thousands of Erica on this machine and the trainer lets you know there could be multiple fixes for every air only similar solutions are written down the rest of our darnedest to learn on the job.

Tara Meissinger: The rest we're going to have to learn on the job. Sounds familiar, right? Years ago, when I started fixing these machines, it took me three to six months to feel comfortable finding solutions for even the most basic recurring issues.

Speaker Change: Sounds familiar eight years ago, when I started fixing these machines. It took me three to six months to feel comfortable.

Speaker Change: Do you feel comfortable finding solutions for even the most basic recurring issues. So I can tell you from experience. This is completely overwhelming now that I'm in leadership. The question for me becomes how can I speed up this learning process, how can I empower an empty to feel confident in such an overwhelming situation and ultimately set them up for success.

Tara Meissinger: So, from experience, this is completely overwhelming. Now that I'm in leadership, the question for me becomes, how can I speed up this learning process? How can I empower a MT to feel confident in such an overwhelming situation and ultimately set them up for success at our company?

Speaker Change: At our company.

Tara Meissinger: Ask Adam, or Adam, as we call it, uses AIP in our ontology to serve as a co-pilot for the technician. Adam can take a maintenance work order and link it to a wealth of information, both official and unofficial, and make suggestions about what an MT should try to resolve a problem. So let's try to troubleshoot this machine. As we walk around, we can see that there's a flashing red sensor indicating that there's a cause for the error.

Ask Adam or Adam as we call. It uses AIP in our ontology to serve as a co pilot for the technician.

Speaker Change: Adam can take a maintenance work order and link it to a wealth of information both official and unofficial and make suggestions about what an empty you should try to resolve the problem.

So let's try to troubleshoot this machine as we walk around we can see that there is a flashing red sensor, indicating that it's a cause for the air.

Tara Meissinger: Adam will then take that sensor data and start building out its recommendation. Before Adam, an MT would have to search a variety of databases and manually piece together relevant information, which is a huge waste of time when you have a downed machine right in front of you. Now, at their fingertips, MTs can see work that was previously done on this winder, the spare parts that were replaced, and the results of any quality investigations that took place recently.

Speaker Change: Adam will then take that sensor data and start building out its recommendations before Adam and empty you would have to search a variety of databases and manually piece together relevant information, which is a huge waste of time when you have a down machine right in front of you.

Speaker Change: Now at their fingertips MTS can see work that was previously done on this wind or the spare parts that were replaced.

Speaker Change: And the results of any quality investigations that took place recently.

Tara Meissinger: We have over a million tickets in our system full of tribal knowledge. Adam can read all of these comments that MTs put in and summarize them in the order from easiest to hardest suggestions for how a technician should proceed. Finally, Adam can search our library of controlled documents and only surface relevant instructions for the problem at hand.

Speaker Change: We have over 1 million tickets in our system full of tribal knowledge.

Speaker Change: Adam can read all of these comments that mt's put in and summarizing the order from easiest to hardest suggestions for how a technician should press forward finally, Adam can search our library of control documents and only surface relevant instructions for the problem at hand, So you can.

Tara Meissinger: So you can see here. We just learned from Adam that all we had to do was replace that switch, and now our machine is back to making batteries. As a last step, the technician will record their fix, and it is indexed back within the ADAMS system, enabling the platform to become smarter and more robust as time goes on. Ask Adam is driving down time on our equipment by providing information to empower technicians to get to the correct solution as quickly as possible.

Speaker Change: See here, we just learned from Adam the only had to do is replace that switch and now our machine is back to making batteries.

Speaker Change: As the last step the technician will record their fix and it is indexed back within the atom system, enabling the platform to become smarter and more robust as time goes on.

Speaker Change: Ask Adam's driving uptime on our equipment by providing information to empower technicians to get to the correct solution as quickly as possible. It also serves as a vital training tool for new or even veteran technicians, who are confused on how to proceed with troubleshooting cutting that three to six months learning curve that I mentioned.

Speaker Change: Earlier down to just a few weeks.

Tara Meissinger: It also serves as a vital training tool for our new or even veteran technicians who are confused about how to proceed with troubleshooting, cutting that three to six month learning curve that I mentioned earlier down to just a few weeks. Now, what I showed in this demo is just the start. Everything we do in Version 1 will enable us to build out more features in Version 2, where we're talking about features like smart PM scheduling, which will enable us to go from a preventative maintenance to a proactive maintenance state and start to answer the questions not only what's going to break today but what's going to break tomorrow, and what's going to hold us back in a week or even a month.

Now what I showed in this demo is just the start from Adam.

Speaker Change: Everything we do in version one will enable us to build out more features in version two <unk>. What we're talking about features like smart PM scheduling, which will enable us to go from a preventative maintenance to a proactive maintenance state and start to answer the questions not only whats going to break today, but what's going to breakdown.

Speaker Change: What's going to hold us back in a week or even a month.

Tara Meissinger: We're also considering automated spar parts management to be able to feed the technicians the parts that they need when they need them at a low cost. Looking even further into the future, with our continuing partnership with Palantir, we're building ADAM as a gift for our future sister plant in DeSoto, Kansas.

Speaker Change: We're also considering.

Speaker Change: Automated spare parts management to be able to feed the technicians the parts that they need when they need them at a low cost.

Speaker Change: Looking even further into the future with our continuing partnership with pollen tier we are building Adam as a gift for future sister plant in Desoto, Kansas with Adam The Kansas <unk> will also be able to tap into all of the experience and tribal knowledge that we've gained over here in Nevada and deploy immediately halfway across the country.

Tara Meissinger: With ADAM, the Kansas MTs will also be able to tap into all the experience and tribal knowledge that we've gained over here in Nevada and deploy it immediately halfway across the country. I want to thank the Palantir team again for making all of this possible for us. I've been turning over this idea in my mind for years, but to see it come to life quickly and positively affect my technicians has been more rewarding than I can say.

Speaker Change: <unk>.

Speaker Change: I want to thank the talented <unk> team again for making all of this possible for us I've been turning over this idea in my mind for years, but to see it come to life quickly and positively affect my technicians has been more rewarding than I can say this has truly been one of the most extraordinary projects of my career and I'm very grateful to be invited here.

Tara Meissinger: This has truly been one of the most extraordinary projects of my career, and I'm very grateful to be invited here to share it with you all today. Thank you. Please welcome the winner of the Fastest AI-Driven Business Model Pivot from Caz Investments, James Stewart. Good morning.

Sure to share it with you all today. Thank you.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Please welcome the winner as fastest AI driven business model pivot from Kaz investments James Stewart.

James Stewart: Thank you.

James Stewart: Okay.

James Stewart: [music].

James Stewart: Good morning, My name is James Stewart, and the Vice President of Finance at <unk> investments.

James Stewart: My name is James Stewart, and I'm the Vice President of Finance at Caz Investments. Caz Investments provides exclusive access to unique alternative investments. We've experienced a tremendous amount of growth over the past few years, and this has been supercharged over the past couple months. In February, our chairman, Christopher Zook, released a new book with Tony Robbins called The Holy Grail of Investing. We have also launched a new fund called the Strategic Opportunities Fund that dramatically reduces the minimum investment required by individual investors.

James Stewart: Cash investments provides exclusive access to unique alternative investments.

James Stewart: We've experienced a tremendous amount of growth over the past few years and this has been supercharge over the past couple of months.

James Stewart: In February our chairman Christopher <unk> released a new book with Tony Robbins called the Holy Grail of investing.

We also launched a new fund called the strategic opportunities fund that dramatically reduces the minimum investment required by individual investors.

James Stewart: Our decision to partner with an AI firm has brought about a plethora of opportunities for us. We are so grateful to AIP for being able to build an online lead triaging system. We partnered with them six months ago, and I want to share with you what we've been able to accomplish and how we did it with AIP. We can now process over 100 times more leads with the same amount of resources. This has reduced our overall lead processing time by over 90%.

James Stewart: Our decision to partner with an AI firm has brought about a plethora of opportunities for us.

James Stewart: We are so grateful to AIP to be able to build an online lead triaging system.

James Stewart: And we partner with them six months ago, and I want to share with you what we've been able to accomplish and how.

James Stewart: With AIP.

James Stewart: We can now process over 100 times more leads with the same amount of resources.

James Stewart: This has reduced our overall lead processing time by over 90%.

James Stewart: In just two months, we have managed 7,000 leads. This would not have been possible without AI. Together, we've automated 85% of our workflow while leaving humans at key checkpoints. These humans in the loop provide feedback to make the system smarter and more efficient. AIP is able to extract key information from unstructured data sources such as emails, enrich that data, and send intelligent communications both internally and externally. And all of this works together seamlessly with our current systems, such as Salesforce. But not only has AIP made us more efficient, we have been more effective. We now have complete visibility over our lead to partner conversion process.

James Stewart: In just two months, we have managed 7000 leads this one not impossible without AIP.

James Stewart: Together, we've automated 85% of our workflow.

James Stewart: While leaving humans at key checkpoints. These.

James Stewart: These humans in the loop provide feedback to make the system smarter and more efficient.

James Stewart: AIP is able to extract key information from unstructured data sources such as emails.

James Stewart: Enrich that data and intelligent communications, both internally and externally.

And all of this works together seamlessly with art.

James Stewart: Current systems such as Salesforce.

But not only has AIP made us more efficient we have been more effective we now have complete visibility over our lead to partner conversion process.

James Stewart: This allows us to tailor the experience for each lead. Utilizing large language models, AIP is able to enrich lead and manager data, proposing a pairing that provides us with the greatest opportunity for a successful conversion. AIP also provides us with key metrics such as manager performance, communication detail, and Financial Analysis.

James Stewart: This allows us to tailor the experience for each lead.

James Stewart: Utilizing large language models AIP is able to enrich lead and manager data proposing a pairing that provides us with the greatest opportunity for a successful conversion.

James Stewart: AIP also provides us with key metrics such as manager performance communication detail.

James Stewart: And financial analysis.

James Stewart: All of this has worked together to help CASA Investments continue to learn, win, and provide the best experience for our partners. I invite you all to come to our demonstration booth later today, where I will go through this process in more detail. Palantir has been an amazing partner, and we continue to look forward to working with them on future projects. For example, we're looking to utilize AIP in our research as we review and analyze future investment opportunities. I hope to see you all there. Thank you.

James Stewart: All of this has worked together to help CASM investments continuing to learn win and provide the best experience to our partners.

I invite you all to come later today to our demonstration Booth I will go through this process in more detail.

James Stewart: Talents here has been an amazing partner and we continue to look forward working with them on future projects.

James Stewart: For example, we're looking to utilize AIP and our research as we review and analyze future investment opportunities.

Speaker Change: I hope to see you all there thank you.

James Stewart: Please welcome the winner of the Most Effective AI Implementation Team from Cummins, Jim Jacob. Hey everybody, it's so good to be here. It's wonderful to see the team here because I came a few years ago.

Speaker Change: Thank you.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Please welcome the winner for most effective AI implementation team from comments Ken Jacobs.

Speaker Change: Yes.

Speaker Change: Yes.

Yes.

Ken Jacobs: Hey, everybody. So good to be here, it's wonderful to see the team here because I came a few years ago. It was not as big as a.

Jim Jacob: It was not as big as he said it was. It's fantastic to see so many people in a cutting-edge era participating in this. So I'm from Cummins, as you heard. A big shout out, obviously, to my team, who's also online watching right now.

Ken Jacobs: As he said, it's fantastic to see so many people in a cutting edge era participating in this so I'm from Cummins as you heard a big Shout out obviously to my team was also online watching right now so a great job guys for the award that we got thank you palen tearful, what you've done and done continue to do.

Jim Jacob: So great job, guys, for the award that we got. Thank you, Palantir, for what you've done and will continue to do, of course, in terms of recognizing the work that we did. I have a couple of things that I want to share with you in terms of how we were able to get to this point. Now, I'm a little biased, maybe.

Speaker Change: Of course.

Speaker Change: In terms of recognizing the work that we did I have a couple of things that I want to share with you in terms of how we were able to get to the success point out I'm, a little biased maybe I've been using patented for about seven years and so the second company M. Starting to use volunteer with and are starting to see great results. So with that let me go through a couple of things.

Jim Jacob: I've been using Palantir for about seven years now, and this is the second company I'm starting to use Palantir with, and I'm starting to see great results. So with that, let me go through a couple of things. With the right players and the right tools, it might sound very simple, but when you do this over and over and over again and realize that, you know, you're not getting too far with some of the technology and with the balance, it's very essential that you put that together.

Jim Jacob: And in many cases, and we've done so many incoming so many projects, but in many cases, once that balance strikes and you have the right technology, it works like crazy. So with that, these are a couple of points that are kind of guiding principles, so to say, getting the right team balance, getting a subject matter expert from the business where the where the problem resides is critical. You probably heard this situation happen in many cases where you'll get the business side of the function side saying, give me the data. I'll figure out what to do.

Speaker Change: Balance.

Speaker Change: With the right players and the right tools it might sound very simple, but when you do this over and over and over again and realize that you're not getting too far with some of the technology and with the balance its very essential that you put that together and in many other cases, we've done so many incoming so many projects but.

In many of the cases once that balanced strikes and you have the right technology it looks like crazy so with that.

Speaker Change: These are couple of points that are kind of guiding principles. So to say getting the right team balance getting a subject matter expert from the business, where this where the problem resides is critical.

Jim Jacob: You give them the data, and then they'll tell you to give me one more column or give me one more piece of information, and now you have multiple problems. So the problem that was just one problem is now multiple problems. Now you have a second problem because now you're interacting with the person who's trying to build a solution for you, and the original business problems are still there, too. So now you have three problems to resolve. So this is not true. I mean, it happens all the time.

Speaker Change: You probably heard this situation happen in many many cases, where you will get the business side of the function side, saying give me the data and figure out what to do you give them that hit.

Speaker Change: And then the hotel you give me one more column give me one more piece of information and now you have multiple problems that are the problem that was one problem now you have a second problem because now you're interacting with.

The person who is trying to bring a solution for you and the original business problems still there too. So now you've got three problems to resolve.

Speaker Change: This is not I mean, it happens all the time.

Speaker Change: We want to run away from that situation, but you want an SME, who is who has the trust from the business side or the function side who's working directly with you and he or she is the center of the solution build process. That's what we were successful in doing it in Cummins with the.

Speaker Change: Multiple opportunities that now the three things that we did very effectively was connecting the information you've heard this over and over again, how important it is for a digit non digital native like Cummins is.

Speaker Change: Youre going to have data everywhere and to bring that data in you've got to have a very effective platform. This goes back to what I told you before if you don't have the right partner youre going to be spending so with volunteer with foundry. It was much easier to make the connections work and connections work not just for one instance, multiple instances.

Speaker Change: In Cummins, we created a platform called the Digital Core, where each of these connections that we made, we put into a digital twin kind of digital core into that repository, and we were able to now refer to that for various problems that were coming up. And that was a big deal. Because once you do that once, you get a result because you're solving for one problem. When you do it the second time, you're going to get exponential results, and so on.

Speaker Change: In Cummins we created.

Platform called the digital core where each of these connections that we made we put it into a digital twin kind of digital core into that reports the trend we were able to now refer that for various problems that were coming up and that was a big deal because once you do that once you get a result, because you are.

Speaker Change: Solving for one problem when you do the second time youre going to get exponential results and so on but you've got to have the right technology to be able to access it fast efficiently smart DNR. That's the difference between foundry as well as with the building ontologies that works really well so with that one big problem that would be how many you might.

Speaker Change: But you've got to have the right technology to be able to access it fast, efficiently, smartly, and all that. That's the difference between Foundry as well as with building ontologies that work really well. And massive gen sets for power generation.

Speaker Change: You could relate to this I'm sure.

Speaker Change: This is supply risk mitigation, where suppliers could create a massive problem for you right through right to earn sales and in our case Cummins is a company that builds massive engines and massive gen sets for power generation. So there are tons of little parts that goes.

Speaker Change: Entities and so you've got to have a fantastic view fantastic hold on what the suppliers. So we will just do it.

Speaker Change: So there are tons of little parts that go into these. And so you've got to have a fantastic view, and a fantastic hold on what the supply is. We built a platform with Foundry and, of course, with Ontologies to be able to drive that right through, take the risk out so that you had good visibility, right? So this project, the business sponsor of it, because we had a subject matter expert work with us, the business sponsor's line was, and it was a great line to hear, "I've not seen this before as a solution."

Speaker Change: Platform with foundry and of course with ontologies to be able to drive that right through take the risk out. So that you had good visibility writers of this project.

Speaker Change: The business is one of it because we had a subject matter expert work with us the business sponsors line more than it was a great line to here I've not seen this before as a solution.

Speaker Change: And that's great to hear because once you do that, that means the business sponsor is going to invest again to do even more with you. So that's a big deal. Second one, I mean, for us, this is a big deal.

Speaker Change: Great to hear because once you do that that means the business sponsors are going to invest again to do even more with you. So that's a big deal second one I mean for US. This is a big deal. This is repair time reduction so what we want to do is keep those trucks on the road as much as you can because the customer wants that to happen. If it's inside the workshop doesn't work very well for neither party and if it's a war.

Speaker Change: This is repair time reduction. So what we want to do is keep those trucks on the road as much as we can, because the customer wants that to happen. So what we have done is we were able to take years and years of repair history, and then we are able to use augmented reality and, of course, AI to be able to figure out the new code that comes in, what repair is required for it. So by the time it comes to the workshop, you're ready with the repair, and it's out.

Speaker Change: Guarantee problem, it's even even worse, so because we are paying for warranty as it as it is in the shop. So what we've done is <unk> been able to go and take.

Speaker Change: Use in years of repair history, and then be able to use.

Speaker Change: Matt augmented reality and then of course.

Speaker Change: AI to be able to figure out the new code that comes in what's the repair that is required for it. So by the time it comes to the workshop, you're ready with the repair and its out so prognostics working really well there again with the balance of these two platforms. So just to quickly summarize what we were able to deal with this digital platform was to be.

Speaker Change: So prognostics worked really well there, again, with the balance of these two platforms. So just to quickly summarize, what we were able to do with this digital platform was to be able to bring all these data sets as we solve problems. And as each of these gets solved, creating a data pipeline, creating the visibility so that the subject matter expert can look at it and then get creative about new problems they can solve, and then be able to create this interoperability with these data sets. And that's the magic.

Speaker Change: To bring all of these datasets as these solve for problems.

Speaker Change: Each of these get sold creating a data pipeline, creating the visibility for that subject matter expert can look at it and then get creative about new problems that can solve and then.

Speaker Change: To create this inter operative interoperability for with these data sets and that's that's the magic I mean connect indifferent enact that's the way we book so with that I mean, if there is more questions on details on what we did more than happy to answer them, but for now that's all we have done for the beginning we have a long way to go with it.

Speaker Change: I mean, connect, interpret, and act. That's the way we work. So with that, I mean, if there are more questions on details about what we did, I'll be more than happy to answer them. But for now, that's all we have done for the beginning.

Speaker Change: We have a long way to go with the relationship we have with Palantir. And then, for this year, we probably have another 10, 15 projects that we would like to take on and do. And each one we do, like I said, that digital core gets even stronger. So with that, thanks a lot, Palantir, for the opportunity to be here. And thanks a lot to the Cummins team.

Speaker Change: The relationship we have with Pan out here and then for this year, we probably have another probably 10 15 projects that we would like to take on and do in each one we do like I said that digital code gets even stronger so with that thanks, a lot volunteer for the opportunity to be here and thanks, a lot the Cummins team great job guys good to be here.

Speaker Change: Great job, guys. Good to be here. Please welcome the winner of the Best AI-Enabled M&A Strategy from MacDermid Enthone, a division of ESI, Timothy Gottsik.

Speaker Change: Okay.

Speaker Change: Okay.

No.

Speaker Change: Please welcome the winner for best AI enabled M&A strategy from Mcdermott and done a division of ESI Timothy got sick.

Speaker Change: Okay.

Speaker Change: [music].

Timothy Gottsik: First of all, I'd like to acknowledge the fact that the work I'm going to describe is the product of a lot of really smart colleagues of mine, both within MacDermid Enthone and Element Solutions, as well as at Palantir. And so I'd like to thank them all for making this possible and allowing me to represent them. So let's talk about what we've done. This is not a bizarre situation, but we're a specialty chemicals company.

Timothy: Hello first of all I'd like to acknowledge the fact that the work I'm going to describe is the product of a lot of really smart colleagues of mine both within Mcdermott enthrone element solutions as well as at <unk>. So I'd like to thank them all for making this possible and allowing me to represent them. So let's talk about what we've done.

Timothy: This is not a bizarre situation but.

Timothy: We are a specialty chemicals company, we've grown a great deal by acquisition in recent years, you may not know our name, but our products are used in almost everything you touch if you drove a car flew in a plane used phone use a laptop or chemistry is used to produce all those things to make them reliable and higher performance.

Timothy Gottsik: We have grown a great deal by acquisition in recent years. You may not know our name, but our products are used in almost everything you touch. If you drove a car, flew in a plane, used a phone, used a laptop, our chemistry is used to produce all those things to make them more reliable and perform better. In about 2016, McDermid and Enzone combined, and after much study, were called McDermid-Enzone

Timothy: In about 2016, Mcdermott and phone combined and after much study called Mcdermott inbound.

Timothy Gottsik: And these two companies, McDermott had a U.S. heritage, both were global; McDermott had a U.S. heritage, and Enthone had a German heritage. And in 2017, I transferred within the company and moved to Germany to take over R&D. I am a chemist, which is kind of strange that I'm talking about this, but that's how life goes sometimes.

Timothy: And.

Timothy: These two companies Mcdermott had a U S heritage both for global that Mcdermott has had a U S heritage anthem had a German heritage and in 2017 I transferred within the company and moved to Germany to takeover R&D I am a chemist.

Speaker Change: Which is kind of like strange that I'm talking about this but that's how it life goes sometimes.

Timothy Gottsik: So these two entities combined, and then in 2021, we combined Coventia, and at that point, we sort of broke the camel's back. Our data systems were not aligned, we had lots of different ERPs, we had lots of different master data, and a lot of it was redundant. And my R&D scientists were spending a huge amount of time having to sort through legacy codes to understand is this material the same as that material, is this product the same as that product. It was compromising our effectiveness. And then, just for the last bullet into the dead horse, we bought a German company HSO recently just to compound the problem.

Speaker Change: <unk>.

Speaker Change: So these two entities combined and then in 2021, we combined <unk> and at that point, we sort of broke the camel's back our data systems were not aligned we had lots of different ERP as we had lots of different master data a lot of it was redundant.

And my R&D scientists were spending a huge amount of time, having to sort through legacy codes to understand is this material is the same as that material as this product the same as that product it was.

It was compromising our effectiveness and then just for the last bullet into the dead Horse, We bought German company HSR recently, just to compound the problem. So we had too much data a lot of it was redundant and we need to figure out a way to fix it well.

Timothy Gottsik: So we had too much data, a lot of it was redundant, and we needed to figure out a way to fix it. Well IT, a very helpful, smart group of people, said, ah, we'll just put in, you know, one ERP, SAP S4, all quite easily done, of course, right? Well, as soon as we started studying this and looking at it, this challenge in the middle was a bullseye on my head because we were going to be the rate-limiting step in trying to cleanse and align data, and we were going to end up being the thing that prevented this from happening, which was not going to be pretty.

Speaker Change: Very helpful. Smart group of people said I will just put in one ERP SAP for all quite easily done of course right.

Speaker Change: Well as soon as we started to studying this and looking at it. This challenge in the Middle was a bulls eye on my head because we were going to be the rate limiting step in trying to cleanse and aligned data and we were going to end up being the thing that prevented this from happening, which was not going to be pretty and we needed to find a solution.

Timothy Gottsik: And we needed to find a solution. After one very unfortunate project, we were introduced to Palantir, our CEO, and he said, hey, you should take a look at this solution and see if it's something that we can use. And I almost wanted to retitle my talk, You Had Me at Ontology. Because the first time I saw the presentation, I was like, oh my God, this is a gift from the heavens. And fortunately, it turned out to be worth all the hype that was there.

Speaker Change: After one very unfortunate project, we were introduced the volunteers CEO and said Hey, you should take a look at this solution and see if it's something that we can use.

Speaker Change: And I almost wanted to re titled My My talk you had me at ontology because the first time I saw the presentation I was like Oh My God. This is a gift from the heavens.

Speaker Change: Unfortunately, it turned out to be worth.

Speaker Change: Worth all the hype that was there we were able to create the ontology, which was interesting to me, but I already knew is messed up the really cool thing for me was it now at a diagram that showed how badly messed up it was.

Timothy Gottsik: We were able to create the ontology, which was interesting to me, but I already knew it was messed up. The really cool thing for me was that now I had a diagram that showed how badly messed up it was. But they were able to do all the things that I think many of you already realize here, hooking into legacy systems, diagram it out, and giving a really excellent representation of the current situation.

Speaker Change: But they were able to do all the things that I think many of you already realized there where the hook into legacy systems diagram. It out give a really excellent reference representation of the current situation and then we designed some very nice apps in foundry to figure out how to sort through all of it to speed it and we've done a number of different apps in foundry. This is the original.

Timothy Gottsik: And then we designed some very nice apps in Foundry to figure out how to sort through all of it to speed it up. And we've done a number of different apps in Foundry. This is the original one, and it's still underway for cleansing raw material data.

<unk>, one and it's still underway for a cleansing raw material data.

Timothy Gottsik: This is the dashboard that I look at every morning to understand how our progress is going. It shows me how many materials have been cleansed in every one of the ERPs. It contains a data quality marker called the collapse factor, which one of my colleagues, Nadia Billan, amazingly, with like 2% of the available data, predicted that it was going to be around five and a half. That means that we have five and a half redundant codes for every unique raw material, is what that amounts to.

Speaker Change: This is the dashboard that I look at every morning to understand what's our progress. It shows me how many materials have been cleanse and every of the ERP.

It contains a data quality marker called the collapse factor, which one of my colleagues to not have your bill and amazingly with like 2% of the available data predicted that it was going to be around five five that means that we have five and a half redundant codes forever unique raw materials is what that amounts to and it's proved to continue that way and continues to be one of the better predictions I have ever seen.

Timothy Gottsik: And it's proved to continue that way and continues to be one of the better predictions I've ever seen. So we are now in a position where I don't think we're going to be the rate-limiting step on an S4 implementation that we are doing in Germany this year. And we have other modules that we're working on that are going to speed other aspects of this process as well. Even more exciting, potentially, is what we are going to do next in the future.

Speaker Change: So we are now in a position where I don't think we're going to be the rate limiting step on an SBA P. S. Four implementation that we are doing in Germany. This year and we have other modules that were working on that are going to speed other aspects of this process as well.

Speaker Change: Even more exciting potentially is what we're going to do next in the future. So like all of you I attended a boot camp not too long ago, and was able to sit down with one of the <unk> colleagues roofing and in 90 minutes. He built in AIP module that was able to read a very poorly made scan of a supplier data.

Timothy Gottsik: So, like all of you, I attended a boot camp not too long ago and was able to sit down with one of the Palantir colleagues, Vruthik, and in 90 minutes, he built an AIP module that was able to read a very poorly made scan of a supplier data sheet and compare it against the specification of a raw material and issue a remarkably good judgment on whether or not it was equivalent to what we needed it to be This has just sort of, you know, blown our minds, and now we're going through the process of figuring out how we can do this.

Speaker Change: <unk> and compare it against the specification of our raw material and issue a remarkably good judgment on whether or not it was equivalent to what we needed. It to be this is just sort of blown our mind and now we're going through the process of figuring out how do we do this this unlocks the value creation capability of my chemist, who instead of shuffling papers.

Timothy Gottsik: This unlocks the value creation capability of MyChemists, who, instead of shuffling papers and calling vendors, are able to do what they do best, which is to look at chemistry and figure out how we can make better products for the uses that we intend. And it also positions us as a company to be able to process and integrate future acquisitions much, much faster, which is going to let us deliver value to our customers that much quicker. So with that, I thank you all very much for your attention and Palantir for making such a cool product. And that's what my gig is. Please welcome Jack Dobson from Palantir. Good morning. My name is Jack Dobson.

Speaker Change: And calling vendors are able to do what they do best which is to look at chemistry and to figure out how we can make better products for the uses that we intent and it also positions us as a company to be able to process and integrate future acquisitions much much faster, which is going to let us deliver value to our customers that much quicker.

With that I. Thank you all very much for your attention and Pal interior for making such a cool product.

Speaker Change: That's my gig.

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Please welcome from talent here Jack Dobson.

Speaker Change: Okay.

Jack Dobson: Good morning, My name is Jay Dobson on one of our commercial AIP leads his talents here when we started running AIP boot camps. In 2023, we sold just how quickly we could take out customers from their first experience of AI to have deployed use case in most cases in a matter of hours and days.

Jack Dobson: I'm one of our commercial AI leads here at Palantir. When we started running AI boot camps in 2023, we saw just how quickly we could take our customers from their first experience of AI to a deployed use case, in most cases, in a matter of hours and days. And as we near 850 boot camps in the U.S. and around the world, we've seen the value of bringing those customers together to share and collaborate, learning from each other's solutions, developing their intuition, and forwarding their operational AI journeys.

Jack Dobson: And as we near 850 boot camps in the U S and around the world, we've seen the value of bringing those customers together to share and collaborate learning from each other solutions developing their intuition and forwarding that operational AI journeys.

Jack Dobson: Our individual customer boot camps have quickly become cross-industry showcases, with customers going from their first hands-on experience with AI to demoing their first end-to-end workflow. As our customers go from their second to their third to their fourth boot camp, many of whom are in this room today, we've seen how quickly they build the intuition and the experience to outpace even us, going from their first hands-on experience with AI to a fully- And quickly, the question emerged from our customers: we have so many ideas, and we want to move faster. How can we accelerate the journey from bootcamp to production?

Jack Dobson: Individual customer boot camps have quickly become cross industry showcases with customers going from that first hands on experience of AI to demo ing. The first end to end workflow.

Jack Dobson: As our customers go from the second to the third to the fourth bootcamp many of whom are in this room today, we've seen how quickly they build intuition and the experience to outpace even enough going from their first hands on experience to a fully fledged AI enhanced enterprise and quickly the question emerge.

Jack Dobson: From our customers we have so many ideas and we want to move faster how can we accelerate the journey from boot camp to prod, what can we learn from other policy of users and one better how can we contribute back to that high of mines.

Jack Dobson: What can we learn from other Palantir users? And one better, how can we contribute back to that hive mind? So to meet this demand, we've been building solutions to support our customers' journeys from initial idea to use case, leveraging AIP to take something as simple as a whiteboard of ideas, identify applicable use cases, and then deploy them. I'd like to show you an example of this today. So I'm at an AIP boot camp.

Jack Dobson: So to meet this demand we've been building solutions to support our customers' journeys from initial idea to use case, leveraging AIP to take something as simple as a white board of ideas identify applicable use cases, and then deploy it I'd like to say you are an example of this today.

Jack Dobson: I've been hands-on with the platform, I've built my intuition, and now we're deep in discussion about the potential applications of AIP at my company. Ideas are sprawled across notes and whiteboards, and we deepen that discussion around those different potential use cases. How complex will they be? Where do we start? How long is it going to take?

Jack Dobson: So I'm the AIP bootcamp I've been hands on with the platform I built my intuition and now with deepen discussion about the potential applications of AIP.

Company ideas this broad across notes whiteboards and with deepened that discussion around those different potential use cases, how complex will they be.

Jack Dobson: Where do we stop how long is it going to take this aviation is great, but what if we could skip the hypothesizing, what if we could just get hands on with products move past ideas and skip straight to build.

Jack Dobson: This ideation is great, but what if we could skip the hypothesizing? What if we could just get hands-on with products, move past ideas, and skip straight to build? So I'm going to drop my ideas and my notes, in whatever form, into AIP to begin the process of turning them into use cases. AIP logic will facilitate the handoff between multiple LLMs to work on any structured or unstructured data that I've

Jack Dobson: So I'm going to drop by ideas in my notes in whatever form into AIP to begin the process of turning them into use cases, AIP logic will facilitate the handoff between multiple LMS to work on any structured or unstructured data that I provided in this case, a vision based model extract c&c's from a picture.

Jack Dobson: And in this case, a vision-based model extracts the entities from a picture of the whiteboard at this boot camp and starts mapping these to a repository of products. This enables us to draw on the Palantir Builder community, one that you're all going to experience here today at AIP.com. This community is solving complex, impactful operational challenges and choosing to feed their solutions back into the product, becoming an asset to all of you. So why reinvent the wheel when you can learn from others? and why just learn when you can build?

Jack Dobson: The whiteboard at this boot camp and starts mapping these through a repository of products. This enables us to draw on the palliative builder community one that you're all going to experience here today at AIP com. This.

Jack Dobson: This community is solving complex impactful operational challenges and choosing to feed their solutions back into product, becoming an asset to all of you.

Jack Dobson: So why reinvent the wheel when you can learn from others and why just learn when you can build.

Jack Dobson: Sure.

Jack Dobson: So with potential use cases identified from my boot camp, this is a map against Palantir's marketplace to identify available products linked to notionalized case studies of similar deployments across industries, bringing us right up to today. Every keynote and public demo is easily linked to from each use case, and this is a rich network of ideas and potential value, all stemming from that one whiteboard. But we don't stop there.

Jack Dobson: So with potential use cases identified for my <unk>. This is mats against balances marketplace to identify available products linked to nationalize case studies of similar deployments across the industries, bringing us write off today every keno and public demo as easily linked to from each use case and this is a rich net.

Jack Dobson: Work of ideas and potential value all stemming from that one whiteboard.

Jack Dobson: But we don't stop there once I've identified by use case I want to move from learning to building.

Jack Dobson: Once I've identified my use case, I want to move from learning to building. AIP can generate a solution design for me on the spot, mapping this newfound context to my company, my enterprise, my ontology, and my existing data connection. And if I'm happy with this, we can jump straight into a deployable version of this workflow. With one click, I can jump into the user apps and start exploring a virtual version of this product. This deployment package bundles up everything I need to get started.

<unk> can generate a solution designed for me on the spot mapping. This new found context to my company My Enterprise My ontology and my existing data connections.

Jack Dobson: And if I'm happy with this we can jump straight into a deployable version of this workflow with one click I can jump into the user apps and start exploring a notional version of this product.

Jack Dobson: This deployment packages up everything I need to get started a notional ontology AIP logic automate and front end user applications I can explore build my intuition.

Jack Dobson: A notional ontology, AIP logic, automation, and front-end user applications. I can explore, build my intuition, and actually start testing the solution live. This is now an asset that I can start editing as well.

Jack Dobson: Actually start testing the solution live.

Jack Dobson: This is now an asset that I can start at the thing as well I can start tweaking the ontology feeding my own data in editing the logic, making it cost them to my enterprise building on this foundation gives me everything I need to start moving fast and getting to production quicker faster than ever before.

Jack Dobson: I can start tweaking the ontology, feeding my own data in, editing the logic, making it custom to my enterprise. Building on this foundation gives me everything that I need to start moving fast and getting to production quicker, faster than ever before. This is going from an idea to a use case in a matter of seconds. And this is just one example. A multitude of use cases are already available on this marketplace, and I'm very excited to announce that, exclusively for AIPCon attendees, workflows inspired by every keynote you've just seen on this stage are available to deploy to your Palantir instance today.

Jack Dobson: This is going from an idea to a use case in a matter of seconds.

Jack Dobson: And this is just one example.

A multitude of use cases are already available on this market place and I am very excited to announce that exclusively for AIP con attendees workflows inspired by every keynote you've just seen on this stage are available to deploy <unk> instance today.

Jack Dobson: So, to bring us up to date, everyone in this room is part of that community, having booted up and deployed AIP within your organizations, and AIPcon is enabling more collaboration than ever before. You've already seen the incredible work of eight keynotes on this stage, and after lunch, more than 60 customers and partners are going to be presenting at the various sessions throughout the afternoon. You'll be able to get hands-on at the Demo Expos, with live presentations by our customers of their AIP workflows, including Northwind, Trinity, CSX, AMGI Studios, 8BC, Partstown, and more.

Jack Dobson: Yes.

Jack Dobson: Okay.

Jack Dobson: Yes.

Jack Dobson: Sure.

Jack Dobson: So to bring us up to date, everyone. In this room is part of that community, having boot camps and deployed AIP within your organizations and AIP corn is enabling more collaboration than ever before you've already seen the incredible work of eight keynotes on the stage and after lunch more than 60 customers.

Jack Dobson: <unk> are going to be presenting at the various sessions throughout the afternoon.

Jack Dobson: You'll be able to get hands on the demo exposed with live presentations by our customers of the AIP workflows, including North wind Trinity see effects Amg's Studios ABC pottstown more you'll meet the masters the pinnacles of the policy community sharing their experiences tackling some of our toughest <unk>.

Jack Dobson: You'll meet the masters, the pinnacles of the Palantir community, sharing their experiences tackling some of our toughest problems and sharing new ideas with this community. That includes Aramark, Cone Health, HDR, Jacobs, Kinder Morgan, Little Otter, and many more.

Jack Dobson: <unk> and sharing new ideas with this community.

Jack Dobson: Aramark cone health HDR Jacobs, Kinder Morgan Little Ulta and many more.

Jack Dobson: And then finally, of course, you'll apply that newly established intuition in RapidFire Bootcamp. But these aren't your traditional bootcamps led by Palantir. These are customer and partner-led bootcamps led by some of the flagship builders in our community. That's BCG, Guardian Premier Solutions, United Airlines, Moody's, Beyond Meat, Intuit, OpenAI, and others. An entire day built and led by you, given back to this community.

Jack Dobson: And then finally of course, you'll apply that newly established intuition and rapid fire boot camps.

But these aren't your traditional boot camps led by talented these are customer and partner led boot camps led by some of the flagship builders in our community that BCG Guardian Premier solutions, United Airlines, Moody's beyond meat into at open AI and others.

Jack Dobson: An entire day built and led by you shed back to this community.

Jack Dobson: I want to end by highlighting how AIP is further enriching your experience today. The entirety of AIP.com is modeled in AIP, which is being used to enrich and manage all of the operations of AIP.com Dynamic scheduling, attendee management, waitlist, and so much more. And with the content of all of your presentations, workflows, boot camps, and beyond, we can enable far richer connections than ever before. AIP Logic has been used to match content to every attendee, taking into account your past boot This, in turn, is operationalized by informing dynamic scheduling, which is updating your agendas in real time throughout the day.

Jack Dobson: I want to end by highlighting how AIP is further enriching your experience today.

Jack Dobson: The entirety of AIP column is muddled in AIP, which is being used to enrich our managed all of the operations of AIP called dynamic scheduling attendee management waitlist and say much more.

Jack Dobson: Content of all of your presentations workflows boot camps and beyond we can enable far richer connections than ever before.

Jack Dobson: AIP logic is being used to match content to every attendee taking into account youll path boot camps real business challenges recommended content that is unique to each and every one of you.

Jack Dobson: This in turn is operationalized by informing dynamic scheduling, which is updating your agendas in real time throughout the day.

Jack Dobson: And you've also been prompted to communicate with us. You can text AIPCON and actually send ideas, questions, and thoughts. These are all being linked back automatically to your private attendee profile, allowing us to further enrich our collaboration with you in real time throughout the event, as well as after today as well.

Jack Dobson: And you've also been prompted to communicate with US you can text AIP con and actually send ideas questions thoughts. These are all being linked back automatically to your private attendee profile, allowing us to further enrich our collaboration with you real time throughout the event as well as after today as well.

Jack Dobson: This is the Palantir community powered by AIP. We are so excited to have you here at AIP.com with us. And with that, time to start building. Thank you so much.

Jack Dobson: This is the policy community powered by AIP.

Speaker Change: We are so excited to have you here AIP come with us.

Speaker Change: <unk>.

Speaker Change: Non stop building.

Speaker Change: So much.

Speaker Change: Sure.

Speaker Change: Okay.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Thanks.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: [music].

Speaker Change: Okay.

Speaker Change: Okay.

Speaker Change: Yeah.

Speaker Change: Yeah.

Speaker Change: Yes.

Speaker Change: [music]. Please welcome from Palin Chair Sasha Spivak.

Jack Dobson: Please welcome from Palantir, Sasha Spivak. Hi, everybody. Welcome back to AIPcon. We were here just three months ago, launching an entirely new product and putting it into the hands of our customers for the first time. And since then, we've been busy.

Unknown Attendee: Hi, everybody welcome back to AIP Cotton, we were here just three months ago, launching an entirely new product and putting it into the hands of our customers for the first time and since then we've been busy but more importantly, our customers have been busy we built our business alongside customers who need to do real.

Jack Dobson: But more importantly, our customers have been busy. We built our business alongside customers who need to do real work, and they need to do it fast. Our customers don't have 12 weeks to spend on proofs of concept.

Unknown Attendee: Work and they need to do it fast our customers don't have 12 week to spend on proofs of concept. They don't have a year to spend ahead of that on scoping that proof of concept and they certainly don't have two years ahead of that to spend with legal on papering. The proof of concept our customers need to go from zero to use case.

Jack Dobson: They don't have a year to spend ahead of that on scoping that proof of concept, and they certainly don't have two years ahead of that to spend with legal on papering the proof of concept. Our customers need to go from zero to use case inside of five days, and our customers want to show what they're doing, not talk about it. And so today, we're so excited to share the stage with customers who are doing just that.

Unknown Attendee: Inside of five days and our customers want to show what Theyre doing not talk about it and so today. We're so excited to share the stage with customers who are doing just that we have more than 180 unique organizations represented in this room alone we have more than 40 speakers across this morning's keynotes and this afternoon's warm up.

Jack Dobson: We have more than 180 unique organizations represented in this room alone. We have more than 40 speakers across this morning's keynotes and this afternoon's warm-ups, demo expo, and master classes. We have customers gathering their employees at their own offices and joining us for the live stream. One customer has more than 100 employees in their Chicago office joining us. Another customer has gathered another 700 employees across their offices in Atlanta and Dallas.

Unknown Attendee: And demo Expo and Masterclasses, we've customers gathering their employees at their own offices and joining us for the live stream one customer has more than 100 employees and their Chicago office, joining us another customer has gathered another 700 employees across their offices in Atlanta, and Dallas more than 40.

Jack Dobson: More than 40 organizations have onboarded to our software in the last two weeks alone and another 10 in the last hour in preparation for getting hands-on with the keyboard this afternoon in a series of boot camps with us. That ambition is keeping us hungry, it's keeping us busy, and we're so excited to share that with you today. So with that, I'd like to introduce Alex. Please welcome from Palantir, our CEO and co-founder, Alex Karp. And for $9.99, you can have your teeth drilled.

Unknown Attendee: Organizations have onboard Ed to our software in the last two weeks alone and another 10 in the last hour and preparation for getting hands on keyboards. This afternoon, and a series of boot camps with us.

Unknown Attendee: <unk> ambition is keeping us hungry, it's keeping us busy and we're so excited to share that with you today, so with that I'd like to introduce Alex.

Unknown Attendee: Please welcome from Palin Chair, CEO and co founder Alex Karp.

Alexander C. Karp: It changed the music is a little dental.

Alexander C. Karp: Andrew I'm going to drill your teeth.

Speaker Change: And for 999, you can never teeth drilled.

Alexander C. Karp: I'm sorry to have you here. A lot has happened for us. When we finally had revenue... In about 2008, we started to look for a place that was not a basement.

Speaker Change: Sure.

Alexander C. Karp: Delay to have you here.

Alexander C. Karp: Sure.

Alexander C. Karp: A lot has happened for us when we.

Alexander C. Karp: When we finally had revenue.

Alexander C. Karp: And about 2008, we started to look for a place that was not a basement.

Alexander C. Karp: And in the early days in that basement, we invited our first customer, which was... A famous clandestine service in America, and we had no product, we had no people, so we all kind of gathered around, much like you try to make yourself larger in front of a bear so it doesn't eat you, to prove we had people. And that's how it began. Then when we moved over to Palo Alto, it was a big moment for Palantir, and we felt we really did shift the history of the world, but especially in Civil Liberties.

Alexander C. Karp: And and in the early days in that basement, we invited our first customer which is a.

Alexander C. Karp: Famous clandestine service in America, and we had no product we had no people. So we all kind of gathered around much like you tried to make yourself larger in front of a bear so it doesn't need.

Alexander C. Karp: To prove we had people and.

Alexander C. Karp: And that's how it began and then we when we moved over to Palo Alto. It was a big moment for Palmateer.

Alexander C. Karp: And Ah.

Alexander C. Karp: Kind of we felt we had arrived because we had built this product P G which.

Alexander C. Karp: Really did shift the history of the world, but especially in Europe by reducing terrorism in accordance with.

Alexander C. Karp: And one of the truisms of business is you're really succeeding at the place where no one believes you actually care. So I cared, we cared about data protection and civil liberties, and we went very deep on these issues. And some of those issues, not all of them, are super important for the AIP product that we're delivering. By the way, we launched the product just under five months ago, and now we have 150 companies using that product, 50% growth in the last month, and we're changing our go-to-market strategy because we have no resources away from 12-month pilots to essentially a boot camp where we all roll up our sleeves over two days and like, What better way to do it?

Alexander C. Karp: Civil liberties, and one of them one of the Truisms of businesses, you're really succeeding at the place where no. One believes you actually care. So I cared we cared about data protection civil liberties, and we went very deep on these issues.

Alexander C. Karp: And some of those issues not all of it.

Alexander C. Karp: Our super important for the AIP product that we're delivering by the way. We began we launched the product just under five months ago and now we have 150 companies using that product 50% growth from last month.

Alexander C. Karp: And we're changing our go to market because we have no resources away from 12 months pilots too.

Essentially a boot camp, where we all roll up our sleeves over two days and like Great. We'll show you a working.

Alexander C. Karp: But both because quite frankly, we don't have the resources and we like to humiliate people out there, saying their products, so what better way to do it.

Alexander C. Karp: We will show you your enterprise working in two days. Talk to other people. But the central reason we have had these precursor technologies, ontology, branching, and understanding of what it would mean to interact with algorithms, whether they're normal algorithms of the kind that are used on the battlefield or large language models, which are not exact in a way where you can extract value that needs to be exact under conditions where you, for legal, moral, and ethical reasons cannot expose, and business survival reasons, your knowledge of the world to a third party'

Alexander C. Karp: We will show you your enterprise working in two days.

Alexander C. Karp: Talk to other people, but.

Alexander C. Karp: The Central reason we have.

Alexander C. Karp: Had these precursor technologies ontology branching.

Alexander C. Karp: And understanding of what it would mean to interact with algorithms whether they are normal algorithms of the kind that are used on the battlefield are large language models.

Was a deep understanding of MLP.

Alexander C. Karp: How do you ask and answer questions or they're different in this context, but the core issue. We had in building that product was had a unified datasets, where part of the data has to remain unknown to either machine learning or to the end user.

Alexander C. Karp: And those products are not isomorphic those questions are not exactly isomorphic too how do you interact with a large language model, which is not exact in a way where you can extract value that needs to be exact under conditions, where you for legal moral and ethical reasons cannot expose and business survival reasons, you're now.

Alexander C. Karp: <unk> of the world to a third party's M O M.

Alexander C. Karp: There are jarringly similar use cases, and this is why we have all these people adopting the product, because quite frankly, unless you spent 20 years looking at them, you would make certain mistakes. Like the text at the time, because intelligence operatives have to work with things they know are true and data sets that they are hoping are true, and at the time, you really couldn't interact with the text very easily.

Alexander C. Karp: Jarringly similar use case and this is why.

Alexander C. Karp: We have all these people adopting the product because the quite.

Alexander C. Karp: Quite frankly, unless you spent 20 years looking at those you would make certain mistakes.

Alexander C. Karp: Like the the text that at that time.

Alexander C. Karp: With that because intelligence operatives have to work with things they know our true and datasets that they would are hoping our true and at the time, you really couldn't interact with the tax very easily.

Alexander C. Karp: And so dealing with something that's become very, very powerful, but that's not exact, and the kind of exact algorithmic things you would need to make operational decisions to write to your enterprise. It just takes a long time to believe that that's how you have to build it, and longer to build a product. And so I associate Palo Alto with our ascendancy, this building with our ascendancy, and our commercial business and, quite frankly, our government business in the United States with a really cool ride. And so, and then just, an obvious addendum. I was just at this gathering organized by Senator Schumer, and there are many things that happened there, but it's just.

Alexander C. Karp: So.

Alexander C. Karp: <unk> with the shopping has become very very powerful, but thats not exact and the kind of exact algorithm attic things you would need to make operational decisions to write to your enterprise.

Alexander C. Karp: It's just takes a long time to believe that that is how you have to build it in longer to build the product and so I associate Palo Alto with our ascendancy.

Alexander C. Karp: This building with our tendency and our commercial business and quite frankly, our government business in the United States.

Alexander C. Karp: With a really cool right.

Alexander C. Karp: And.

Alexander C. Karp: And.

Alexander C. Karp: And so and then just.

Alexander C. Karp: Obvious addendum.

Alexander C. Karp: I was just at this.

Alexander C. Karp: Gathering organized by.

Alexander C. Karp: Senators leader Schumer, and there are many things that happen there, but it's just.

Alexander C. Karp: It's very hard to explain how, in reality. If you even took the people who were not in the room and then the people who, like, the league below them and the league below them, you would still have a stronger tech community than any other country in the world. It's just actually kind of bonkers.

Alexander C. Karp: It's very hard to explain how.

Alexander C. Karp: In reality.

Alexander C. Karp: If you even took the the people who were not in the room and then the people who.

Alexander C. Karp: Like the league below them in the league below them, you would still have a stronger tech community than any other country in the world.

Alexander C. Karp: It's just actually kind of bonkers like you have any one person on the tech side in that room.

Alexander C. Karp: Like you have any one person on the tech side in that room has a larger tech business, than probably any other country besides America. And so, we're at this moment where, you know, and we're fighting very hard to get adoption on the battlefield, some of which is happening, a lot of it's sensitive, can't really go into it, but obviously we believe not quickly enough. But one of the things that you are doing, quite frankly, without realizing it, is you're putting pressure on our government to adopt AI in a way that will improve all of our lives, with guardrails, in the civil context and with guardrails in the military context, simply because you're going to adopt things and you're going to call, you're going to run into senators and congresspeople and generals and colonels and sergeants and healthcare providers and people at the VM.

Alexander C. Karp: As a larger tech business.

Alexander C. Karp: Than probably any other country Besides America.

Alexander C. Karp: And so we're at this moment where.

Alexander C. Karp: And we're fighting very hard to get adoption on the battlefield.

Alexander C. Karp: Some of which is happening is a lot of it's sensitive can't really go into but obviously, we believe not quickly enough.

Alexander C. Karp: One of the things that you are doing.

Alexander C. Karp: Quite frankly.

Without realizing it is you are putting pressure on our government to adopt.

Alexander C. Karp: AI in a way that will improve all of our lives with guardrails.

Alexander C. Karp: And the civil contacts and with Guardrails in the military context simply because you are going to adopt things and youre going to call you are going to run into Senators and Congress people in generals and Colonels, and charging and health care providers and people at the VA and the like but why is my product better than yours.

Alexander C. Karp: You're like, but why is my product better than yours? And this is probably the most important reason. I mean, there are many reasons why I'm super happy about the work we're doing in U.S. commercials and, quite frankly, the growth is just amazing. I think I never thought we'd get to the point where the growth was so significant. We are just trying to scrape together enough money to figure out how we can do these things. But one of the orthogonal reasons I'm very happy about this is there's no better way to cajole an organization to Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, try to keep me out of purgatory. I'm already skating a fine line.

Alexander C. Karp: And.

Alexander C. Karp: This is probably the most.

Alexander C. Karp: I mean, there are many reasons why I'm Super happy about the work we're doing in U S commercial and quite frankly.

Alexander C. Karp: Just amazing I never thought we'd get to the point, where the growth was so significant we are just trying to scrap to figure out how we can do these things.

Alexander C. Karp: One of them are fogging, our reasons I'm very happy about this is.

Alexander C. Karp: There's no better way to congeal, cajole and organization too.

Alexander C. Karp: Work faster in this case, our clandestine and military services then to say, Okay. Well you know you paid $1 billion you have a jalopy I paid $5 million and I'm running my whole organization.

Alexander C. Karp: And.

Alexander C. Karp: So I'm very passionate about this for lots of reasons, but that's also one reason and I think we have some questions in.

Alexander C. Karp: Try to keep me out of Purgatory have already scaling a fine line Luckily my co founder.

Alexander C. Karp: What are the I didn't appreciate the value of these F shares, which basically means we.

Speaker Change: Uh huh.

Speaker Change: Harder to fire me put dead.

Speaker Change: I go to these meetings and like.

Speaker Change: And like I get away with speaking of plain English because of these F shares, but realizing I'm starting on thin ice so maybe.

Yes.

Speaker Change: So I don't know if someone's going to ask a question.

Alexander C. Karp: Luckily, my co-founder, uh, uh, one of the things I didn't appreciate the value of, we have these F shares, which basically means we, it's very, it's harder to fire me, but then I have no, I go to these meetings and like, I get away with speaking in plain English because of these F-shares, but I'm realizing I'm skirting on thin ice. I don't know, someone's going to ask a question. Please welcome Senior Director of Government Services at Cisco, Mike Yonkers. Morning. Are we going to sit down or stand up? I'm standing, but you do it yourself.

Speaker Change: Please welcome senior director of government services at Cisco, Mike Yonkers morning, or we got to sit down or stand up or outstanding alright.

Mike Yonkers: It's your show. I'm honored to be here with you today. No, no, it's not my show. It's actually your show. You want to stand, hop? I do not want to hop, thank you. I certainly don't want to dance.

Mike Yonkers: You do use your show now honored to be here with NASA that show, it's actually your show and.

Mike Yonkers: You understand hub I do not want to all thank you I certainly don't want it adds.

Mike Yonkers: I used to play the drums, and I used to be able to do that sort of thing, but I can't do that anymore. But I do have a question for you, and you've kind of started talking about that already this morning. I work in the government solutions part of Cisco, and I'm fascinated by this kind of government to commercial, commercial to government. I have some history with In-Q-Tel, and I read Shyam's blog this past weekend about Palantir Government Web Services, and that's what's enabling what we're working for inside of Cisco.

Speaker Change: You're a little [laughter] I used to play the drums and I used to be able to do that sort of thing, but I can't do that anymore, but I do have a question for you and you've kind of started talking through that already this morning.

Speaker Change: I work in the government solutions part of Cisco and I'm fascinated by this kind of government to commercial commercial to government.

Speaker Change: I have some history previously we think you tell and.

Speaker Change: <unk> charms blogs. This past weekend about Pelletier government web services and Thats whats, enabling what were working for inside of Cisco, but the thing I'm fascinated by I'm just fascinated by the company in general that people hear unbelievable and I've watched a lot of companies try to bring technology that was built for the government to commercial and a lot of companies.

Mike Yonkers: But the thing I'm fascinated by, I'm just fascinated by the company in general. The people here are unbelievable. And I've watched a lot of companies try to bring technology that was built for the government to the commercial market, and a lot of companies try to bring commercial technology to the government. And I've seen very few companies succeed, and it feels like you guys are succeeding in doing that. I'm curious, what is it?

Speaker Change: To bring commercial technology to the government and Ive seen very few companies succeed in it feels like you guys are succeeding in doing that I'm curious like what is it what's the secret like how are you able to do this when so many other companies fail at that.

Speaker Change: What's the secret? How are you able to do this when so many other companies fail at it? Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, And we do that by – I mean, it sounds corny, but we are a believer culture. And that believer culture allows us to unify under a belief, which, by the way, does exclude people. Like if you don't think the U.S. government should have the best software in the world, if you're uncomfortable with the fact that we supply our software to the battlefield – in the U.S. and to allies, we respectfully ask you not to join Palantir.

Speaker Change: Unknown Attendee, Keith Weiss, Mariana Mora, Shyam Sankar, John Thornhill, Palantir Tech, it varies day to day, and God knows I've been frustrated with. I get most frustrated by the people I believe in when I feel like it, but

Speaker Change: Well I appreciate the question.

Speaker Change: There is just to be fair somewhat fair to us.

Speaker Change: Or not to make a lot of companies succeeded at this it's just they haven't succeeded in the last 50 years.

Speaker Change: A lot of the innovation I believe one of the reasons America was less divided in the past was that you had tech companies that dedicated themselves to building products.

Speaker Change: Before the government that then repurpose them for commercial.

Speaker Change: So then the Reframing. The question would be like why did that stop and why or why are we able to do it.

Speaker Change: I mean, arguably even succeeding more on commercial then we are anywhere else well first of all we do attract and retain the very best people in the world.

And we do that by.

Speaker Change: I mean, it sounds corny, but we are a believer culture and that believer culture allows us to.

Speaker Change: Unify under a belief, which by the way does exclude people like who.

Speaker Change: You don't think the U S government should have the best software in the world.

Speaker Change: If you are uncomfortable with the fact that we supply our software to the battlefield.

Speaker Change: In the U S into allies.

Speaker Change: We respectfully ask you not to join countered by the way not in like your.

Speaker Change: And idiot.

Speaker Change: We have this belief structure that belief structure allows us to get people. It's like one of these things where the narrowing of the selection leads to a broader ability to deal with different kinds of difficult people.

Speaker Change: You may know from your organizations many of the most talented people in the world have one thing in common they are very difficult.

Speaker Change: And so.

Speaker Change: And then that belief about by the way, obviously I'm not say were not like it varies day to day and God knows I've been frustrated by I get most frustrated by the people I believe in when I feel like but.

Speaker Change: That and then I think the secret is if you are really doing something because you believe in it you are often solving the problem that the customer call. This clandestine surfaces in America or the D O D or somebody it's not the problem they.

Speaker Change: They want to pay you to build its the problem you believe they need to solve and that's a subtle difference in.

Speaker Change: And that's a subtle difference, and giving them a solution, which, unfortunately, in our culture, is often a long services program where the person being paid absorbs no risk, which is one of the reasons why these things often don't work at a cost that's crazy, but the customer is getting what they want. What we're doing is going to a very different level of abstraction and saying, "But what is the question that they are asking that's technical? What will be the question that you should be asking tomorrow, the next day, in a thousand days? And can we build something at that level of abstraction? Don't tell our people that.

Speaker Change: Giving them a solution, which unfortunately in our culture is often a CIT long services program that where the person being paid absorbed no risk, which is one of the reason why these things often don't work at a cost that's crazy.

Speaker Change: But the customer is getting what they want and what we're doing is going to a very different level of distraction in saying, but what is the question that they are asking that's technical what will be the question that is you should be asking tomorrow. The next day and 1000 days and can we build something at that level.

Speaker Change: Extraction.

Speaker Change: The negative version that we have internally is.

Speaker Change: Since we know that in the in the beginning we were very unpopular it's like how good can the product the product has to be good enough that even though we're the least popular people in the room, there is still going to buy it.

Speaker Change: And that's very deepen our DNA like our ugly duckling DNA, it's like a real thing we believe that we're not the ugly ducklings anymore, but don't tell us that certainly don't tell our people that like.

Speaker Change: We need to build products, ugly duckling products, meaning you're going to buy this product. Despite the fact you do not enjoy being in the room with us because it will save your life, that is, And if you build products like that, And then I'll tell you about the commercial thing. I think what is really responding, like if you talk around, talk to people here, you know, of course, we're a commercial company, you're a commercial company, there are always misalignments.

Speaker Change: We need to build products ugly duckling products, meaning you're going to buy this product.

Speaker Change: Despite the fact, you do not enjoy being in the room with us because it will save your life that that is and if you build products like that.

Speaker Change: And then I will tell you on the commercial thing I think we're just really resonating like if you talk around and talk to people here.

Speaker Change: Of course, we're a commercial company you're a commercial company there are always Miss alignments.

Speaker Change: We're much less misaligned with you than I believe any other company you'll ever interact with because, in the end, we want to deliver things that work, because we know that that's how we survive. And then, and then, because we were so bereft of common sense, we have not played by the book, there's a playbook on how you build these companies. And it's like built-in technology, hire thick Salesforce, hire a CEO that Wall Street loves. Okay, our product, that doesn't look like it's not us.

Speaker Change: We're much less misaligned with you than I believe any other company or ever interact with because in the end, we want to deliver things that work.

Speaker Change: We know that that's how we survive.

Speaker Change: And then and then because we were so bereft of common sense. We have not played by the like Theres a playbook on how you build these companies since it's like built in technology higher fixed sales force hire CEO that wall Street loves our.

Speaker Change: Our product.

Speaker Change: It doesn't look that's not us.

Speaker Change: So and.

Speaker Change: So, and so everybody, but that's the playbook you play. Everybody like you go to a venture or any venture, and that's the playbook. So everybody has this one playbook. We have another playbook, and our playbook. It's really hard to get to work.

Speaker Change: So everybody, but that's the playbook you play.

Speaker Change: Everybody like you go to a venture or that's the playbook. So everybody has this one playbook, we have another playbook and our playbook, it's really hard to get to work, but when it does work you end up with.

Speaker Change: But when it does work, you end up with super loyal reference customers and partners, and they run around saying, yeah, okay, the first meeting may be weird, but you'll see. And then, and then America is also shifted. Like we're not that outside the norm anymore, but I think that's how we did it. And for us, um, like. Now I'm very focused on empowering the younger people at Palantir and getting them on the forefront, getting their compensation structure and other things to work for them.

Speaker Change: Super loyal reference customers and partners and they run around saying yeah. Okay.

Speaker Change: Our first meeting may be weird, but.

Speaker Change: You'll see and then and then America has also shifted like we're not that outside the norm anymore, but I think that's how we did it and for us.

Speaker Change: Like.

Speaker Change: Now I'm very focused on empowering the younger people at talent here.

Speaker Change: I like and getting them on the forefront getting their comp structure and other things to work for them.

Speaker Change: Rebuilding the company. It slightly has to be a different company in an LLM AI world, but we must build products that transform the institutions we work with, where we're fully aligned with you succeeding. But like in private, if you taped a conversation with us, what you would find is, sure, we have frustrations with X, Y, just like you do, probably.

Speaker Change: Rebuilding the company in its slightly has to be a different company than an L. M AI world.

Speaker Change: But the we must build products that transform the institutions, we work with where we're fully aligned with you succeeding but like for like in private if you take the conversation with US what you would find is surely a frustration with X Y just like if you do see probably but in a perfect place.

Speaker Change: But you're not going to find conversations where we're like, oh, what is this thin, bullshit piece of product that we can get a salesperson to buy over a steak dinner that they're going to hate? It's not going to work, and it's going to boost our market cap because our CEO is really good at hanging out with analysts. And the analysts believe something they read at Harvard Business School 50 years ago that no one believes except for them.

Speaker Change: Yes, we're not you're not going to find conversations were like Oh, what does this thing bullshit piece of product that we can get a salesperson to buy over steak dinner that theyre going to hate it is not going to work and it's going to boost our market cap because our CEO is really good at hanging out with analysts and analysts believe something they read at Harvard business School 50 years ago that no one believed.

Speaker Change: <unk>, except for them and.

Speaker Change: And, you know, and... And then we don't do that. And if you want that, you go somewhere else. And there's lots of opportunity for that. And that's just not who we are. Thank you. Please welcome Chief Operating Officer at Cone Health, Mandy Eaton. By the way, I apologize for the dental music.

Speaker Change: Yeah, and you know and.

Speaker Change: And then we don't do that and if you want that you go somewhere else and there's lots of opportunity for that and that's just not who we are.

Speaker Change: Well. Thank you you said this in my first conference that I was here and you said iterate with a partner that you can trust and we trust volunteer and it's awesome to work with you guys. So I appreciate the spirit and the answer that question. So thank you. Thank you.

Speaker Change: Okay.

Speaker Change: Please welcome Chief operating officer at Cone Health Mandy Eaton.

Mandy Eaton: But you don't have to sit in the chair. Nice to see you. Unless you want to. If I say, well, you sit, or are you going to stand? Whatever you want. Let's sit. Can I have something to drink?

Mandy Eaton: By the way I apologize for the dental music right now don't have to sit in the chair.

Mandy Eaton: Unless you want to.

Mandy Eaton: Yeah.

Mandy Eaton: If I say, well, you said or he's going to start wherever you want let's say to say did I have something to drink yes.

Mandy Eaton: Yeah. I'm here. Yeah. Well, thank you. I have a couple of questions. One, my first question. Thank you. And thank you for having this conversation. I think this is great.

Mandy Eaton: I'm here, yes. Thank you I have a couple of questions. One my first question. Thank you and thank you.

Mandy Eaton: Our industry, healthcare, typically lags in its adoption of technology. And there are a lot of good reasons for that. I'm curious for those of us who are trying to drive faster adoption. What recommendations do you have for us?

Speaker Change: For having this conversation I think this is great.

Our industry health care typically lags in its adoption of technology and Theres a lot of good reasons for that and curious for those of US who are trying to drive faster adoption what recommendations do you have for us.

Mandy Eaton: We're doing jarringly well in health care, and there's a technical component, and I could go on. There's a lot of product market fit to use jargon. But we again, a lot of our DNA, and I'll get you that, is to survive in an unfriendly environment like for us in the beginning at Palantir. We had to get users first in clandestine services and then in the military. In the Military, these people really know what they're doing.

Speaker Change: Yeah.

Sure.

Speaker Change: We're parenting is doing.

Speaker Change: Currently well in health care, and Theres, a technical filing and I could.

Speaker Change: There's a lot of product market fit to use jargon, but.

Speaker Change: The.

Speaker Change: Again, a lot of our DNA and I'll get to you is like to survive in an unfriendly environment like for us in the beginning at Palmateer, we had to get users Christa clandestine services and then in the military.

Speaker Change: In the military these people are.

Speaker Change: Really know what they're doing the people.

Mandy Eaton: The people, many of them are not particularly technical, and you can supply the best product in the world as a matter of theory. But you have all these issues like, hey, my real problem is not what I'm being told by the corporate people, and I'm going to block adoption unless you solve the problem I care about. And then, but if you identify that problem, you can also go to the people in charge and say, okay, we will get them to adopt this product; we need to build feature sets that get them to use it.

Speaker Change: And many are not particularly technical.

Speaker Change: And you can supply the best product in the world as a matter of theory, but if none of them are using it for.

Speaker Change: First of all it's obviously not valuable and second of all the product can't get better because it gets better and this is before we were doing algorithm attic things at scale really probably gets better by having more users more users mean, the cliche as more users mean more data, but what it really means is more insights on top of the data that allow you to understand the parts of the data that are valuable.

Speaker Change: And now we can do them steroids. So the most important thing forgetting innovation into the hands of the people is getting innovation in the hands of the people and sitting there I think the people running our hospital Division for example have done an amazing job of like you've got to engage with the frontline people what is it that what is your.

Speaker Change: What is your actual problem here by the way.

Speaker Change: A culture of mutual honesty is crucial because.

Speaker Change: Yes.

Speaker Change: In all larger larger organizations, there's always at every level there is alignment, but there's also a misalignment Mike.

Speaker Change: Maybe what the special operator wants to tell you is my biggest impediment to my work as the bone had officer.

Speaker Change: But like how do I build a product where I can do the work and then it feeds up to them, but where theyre not in the way of the work.

Speaker Change: You have all of these issues like Hey, my real problem is not what I'm being told by the corporate people and I'm going to block adoption unless you solve the problem I care about.

Speaker Change: And then but if you identify that problem. You can also go to the people in charge and say, okay. We will get them to adopt this product we need to build feature sets that get them to use it and then because they are using the product you can also guide the people and when I say guidance I, mostly don't mean control like what a lot with them.

Mandy Eaton: And then because they're using the product, you can also guide them. And when I say guide, I mostly don't mean control, like what a lot of the more successful products do. They provide a range. So we did this early on with financial decisions like, if you're working in a health care context, what we'll get broad adoption is like, my dad's a doctor. If I tell my dad, hey, dad, I know how to build a business, he'd be like, well, I read a book. I'll tell you how to do it.

Speaker Change: Our successful products do is they provide a range. So we did this early on with financial decisions like.

Speaker Change: Like if you're working in a health care context, what will get brought adoption is like my dad to a doctor if I tell my dad, Hey, dad, I know how to build a business you'd be like why I read a book I'll tell you how to do it there's something rates of doctors are doctors and.

Mandy Eaton: There's something, right? So doctors are doctors. And but if you provide a range of like, okay, well, this is where we think the decision should go, and you get to iterate, you're empowering the end user. And also, because we don't believe we can survive if you're not using our product. But you have to empower the people on the front end in a way where they want to use it in a way where they still get to make decisions because there's just some things are never like a doctor looks at a patient and will tell you has a feeling. Is this patient healthy or not?

Speaker Change: But if you provide a range of like Okay. Well. This is where we think the decision should go in you get to iterate, you're empowering the end user and that range guidance also gives the people who are managing the P&L of the business a lot of ability to control the kind of costs and risks that they're managing.

Speaker Change: And so what you really.

Speaker Change: The architectural ways in which our product is built for that again, because that's how we learned to survive and also because.

Speaker Change: We don't believe we can survive if youre not using our product, but you have to empower the people on the front end in a way where they want to use it in a way where they still get to make decisions.

Speaker Change: Because there are just some things are never like a doctor looks at a patient and we will tell has a feeling is this patient healthy or not.

Mandy Eaton: And if you're going to tell the doctors that they can't have that, they're not going to use their product. Same thing for nurses. But you can say, okay, we have capacity issues. And if you make these kind of selections, you will no longer have the capacity. By the way, the frustrating part of you having to tell people no all day. And that's how you get adoption.

Speaker Change: And if you're going to tell the doctor that they can't have that theyre not going to use your product same thing for nurses and so but you can say, okay. We have we have capacity issues and if you make these kind of selections view will now no longer have the capacity as you know by the way the frustrating part of you having to tell people know all day.

Speaker Change: And that's how you get adoption and then then you build on top of that and.

Mandy Eaton: And then you build on top of that, and you get a very strong product. That's great. So people are really at the heart of this. I think that leads me to my second question. When we think about AI and similar technology, should we be thinking about them as job replacers, job changers, or just efficiency drivers?

Speaker Change: You get a very strong product.

Speaker Change: Great. So people are really at the heart of this and I think that leads me to my second question is when.

Speaker Change: When we think about AI and similar technology should we be thinking about that is <unk>.

Speaker Change: <unk> replaces job changers or just efficiency drivers.

Mandy Eaton: Should AI and other solutions like this be helping us to keep pace with how fast our information uptake has become, or should we do less or simplify things? I don't think there's one unified answer to this. I think in the health care industry, where you're very constrained by the number of people you could hire anyway, the primary focus will be making people who are already doing a good to very good job much more efficient. So increasing the productivity of the workers. And then there are lots of reasons why.

Speaker Change: Should AI another other solutions like this be helping us to keep pace with how fast our information uptake is has become or can we do less or simplify things.

Speaker Change: I don't think Theres, one unified answer to this.

Speaker Change: I think in the health care industry, what you where you have your very constrained by the number of people you could hire anyway. The primary focus will be making people who are already doing a good to very good job much more efficient.

Speaker Change: So increasing the productivity of the workers you.

Speaker Change: And then there are lots of reasons why so it's like you probably can get you will get a lot more value out of the individual work force you have.

Speaker Change: And that there'll be less replacement than people think but that's also because the delta between what you're being asked to do and what you can do is so large so youre dealing with thin margins heavy regulatory environment and increased demand. So even if we as we increase share efficiency, it's still going to be a.

Mandy Eaton: So even as we increase your efficiency, it's still going to be a long time before that's truly efficient. I think in broader society, what you will find is people who are at the nexus of either high training or high training in regulatory contexts. I do think if you're not trained, you will have to be retrained because the current use of the LLM that's most efficacious is increasing.

Speaker Change: A long time before that's truly efficient.

Speaker Change: The I think in broader society. What you will find is people who are at the nexus of either high trading or high training and regulatory.

Speaker Change: Context, it will augment.

I do think if you're not if you're not trained you will have to be retrained because.

Speaker Change: The current use of the land that's most efficacious as augmenting it still augmenting more than people want to acknowledge it's especially if theres any actual output that matter so like.

Mandy Eaton: It's still augmenting more than people want to acknowledge, especially if there's any actual output that matters. It's like moving from poetry to decision making requires you're augmenting the ability of a trained, thoughtful individual more than you are replacing them in the near term. And so again, I think this is where you'll see the otherwise constrained environments in the U.S. commercial sector outperforming other Western societies because they have the same issues. But they're rejecting the implementation of these technologies either directly or indirectly by making it, you know, most of the great providers are in the U.S. And that's hard for some places to accept.

Speaker Change: It's like moving from poetry two decision.

Speaker Change: Requires you're augmenting the ability of a trained thoughtful individual more than you are replacing them in the near term.

Speaker Change: And so again I think this is where youll see a otherwise constrained environments in the U S commercial.

Speaker Change: Outperforming other western societies, because they have the same issues, but they're rejecting the implementation of these technologies either directly or indirectly by making it most of the great providers are in the U S and that's hard for some places to accept.

Speaker Change: Great. Thank you. Thank you. Great to see you.

Speaker Change: Great. Thank you. Thank you.

Speaker Change: Please welcome from Panasonic Energy, Alex Majewski. Morning, Dr. Karp. Sit, stand. Sitting's good.

Thanks, Steve.

Please welcome from Panasonic Energy Alex Majewski.

Speaker Change: Yeah.

Speaker Change: Martin Dr Carpenter.

Speaker Change: Yes.

Speaker Change: So it's good okay.

Unknown Attendee: Okay. So I want to ask you more of a philosophical question this morning. So, with all the gloom and doom that's in the media these days around AI, especially in the context of arts and culture, what are some of the positive transformations that you think AI is driving humanity towards that most people aren't thinking about?

Unknown Attendee: So I wanted to ask you more of a philosophical question. This morning, so with all the gloom and Doom. That's in the media. These days around AI, especially in the context of Arts and culture. What are some of the positive transformation that you think AI is driving humanity towards that most people aren't thinking about.

Unknown Attendee: Well, I mean, there are some dangers. So, like, I don't – like, there are two camps here, broadly speaking, those that say it's all good and those that say it's all bad. And I think it's kind of pretty good and very good, and then there are – unless it goes all wrong, then it's very bad. Yeah. And – but – First of all, just, you know, this is probably not the right way to answer the question. I always am; my mom's an artist, I grew up around artists.

Speaker Change: Well I mean, there are some danger so like I don't like did there are two camps here in broad broadly speaking those that say, it's all good and those that say, it's all bad and I think it's kind of pretty good and very good and then there are unless it goes all wrong and then it.

Speaker Change: Is there any validation right.

Speaker Change: And.

Speaker Change: But first of all just this is probably not the right way to answer that question I always am my mom's and artists I grew up around artists.

Unknown Attendee: I kind of personally reject, like most people view artists as being painters, photographers, and writers. I honestly, I view building a business as an art form. So, you know, like, I do think that if you change artists to, I want to do a creative endeavor. Creative is defined as, I'm going to build something that is not the way I would be taught in school to build it. However, it deals with the underlying issues of your art.

Speaker Change: I kind of personally reject like most people do.

Speaker Change: You artists as being.

Speaker Change: Painters photographers writers.

Speaker Change: I honestly I view building a business as an art form. So you know like I do think that if you change artists to I want to do in in creative endeavor.

Speaker Change: Creative is defined by I'm going to build something that is not the way I would be taught in school to build it however deals with the underlying issues of of them.

Unknown Attendee: So, like, you do need to know how to be trained to paint in order to be a great artist. And the great painters and writers have almost, extended their art. So if you read philosophy, I could give you a long list, but you can tell every single sentence is from this person.

Speaker Change: Of your art. So like you do need to know how to be trained how to paint in order to be a great artists, but then the great painters and writers are almost all at all of extended their art. So if you read in philosophy I could give you a long list, but you can tell every single sensors from this person.

Unknown Attendee: And if you change it to the basis that almost everything that we enjoy in this world, certainly in this country, was in the beginning some form of art. I think if you approached large language models as they themselves were great artists as opposed to them being great scientists, you'd have a much better, accurate understanding. And therefore, I think they partner very, very well with creative people who also have a very uneven number, of which I would definitely include myself.

Speaker Change: And.

Speaker Change: If you change it to the basis of almost everything that we enjoy in this world certainly in this country was in the beginning some form of art the.

The business that you know.

Speaker Change: You're building the way in which you're adapting the people who create the most value and then you also assume that people who do this are kind of often outside the norm, meaning they don't fit exactly into the current structure as part of the reason they do the art I tend to see the L. M movement.

Speaker Change: <unk>.

Speaker Change: Enormously emancipatory.

Speaker Change: And I also think that.

Speaker Change: My the thing Thats interesting about large language models is there like many artists they are <unk>.

Speaker Change: Absolutely unique in their capabilities and then they also can't really go buy a pack of cigarettes or get a coffee off the ground like they don't add well may hallucinate I.

Speaker Change: I think if you approached large language model says.

Speaker Change: They themselves are a great artists.

Speaker Change: As opposed to their great scientists you'd have a much better accurate understanding and therefore I think they partner very very well with creative people, who also have are very uneven of which I would definitely include myself.

Speaker Change: That's not obvious.

Unknown Attendee: That's not obvious, but it is obvious to everyone who works at Palantir. Believe me. They're like, "Why is Karp bumping his head, turning the wrong left again, telling us to launch this product?". It is one of the reasons why, even personally, I wanted to embrace this earlier than others. And I think it's going to really make people with guardrails, obviously, happier and wealthier. So following up a little bit on what you said about Japanese and American culture and bringing us together. The world seems farther apart than ever in my lifetime. So do you see AI and things like LLM really bringing us back together? No, I'm not making that up.

Speaker Change: But it is every obviously every when it works it pounds you believe me.

Speaker Change: Weis carp pumping has had turning the wrong left again, telling us to launch this product.

Speaker Change: Hum.

Speaker Change: And I think what you will find in our culture again, okay. So we hire a lot of people from France, Interestingly, because France has essentially one school.

Speaker Change: If you go there you get to rural France, and if you don't.

Speaker Change: You don't.

Speaker Change: And so.

Speaker Change: Like you can just see structurally in a country, where everyone gets to go to bat basically because we don't have one school I don't know where you went to school I don't care no one cares.

Speaker Change: It's not even relevant.

Speaker Change: It doesn't matter I don't know what your parents did no one cares literally no one cares it's all the stuff we have to talk about no one cares.

Speaker Change: Oh care about your performance and are you doing something interesting and that's basically it.

Speaker Change: And we can we.

Speaker Change: AI will hypercharge.

Speaker Change: People who are.

Speaker Change: Who are doing something unique and interesting whether it's inside an organization, meaning in a business that exists building new organizations transforming I mean, one of the things I admire about what you guys are doing as you're bringing a hybrid culture of <unk>.

Speaker Change: Japanese production to America.

Speaker Change: That's much much easier, but I would say probably not possible without software and AI.

Speaker Change: But then what's going to make it particularly interesting as it is a hybrid culture, you're bringing certain things that Americans are very pragmatic.

Speaker Change: <unk> attitude.

Speaker Change: Built on top of a way of doing things that is for what youre doing arguably the best in the world, but that isn't that is a AI creation in action.

Speaker Change: You know the.

It really does mean that the value you're creating is.

Pretty much non attenuated from you.

Speaker Change: In a way that it was in the past and it's particularly good if you're sitting in this country, which is why.

Speaker Change: No.

Speaker Change: We see it a pound here people flooding into pound here from all over the world because you can just work in your own contact so.

Speaker Change: I think in the near term for.

Speaker Change: Many many many people that I find inspiring and disproportionately in America.

Speaker Change: This is going to be a freeing technology.

Speaker Change: And especially.

Speaker Change: Especially for people with artistic proclivities.

Right.

Speaker Change: That is one of the reasons why.

Speaker Change: No.

Speaker Change: Even personally.

Speaker Change: I wanted to embrace this.

Speaker Change: Earlier than others.

Speaker Change: And and I think it's going to really.

People with Guardrails, obviously.

Speaker Change: Happier and welfare.

Speaker Change: So following up a little bit on what you said around.

Speaker Change: The Japanese and the American culture, and bring US together the world seems farther apart than ever in my lifetime.

Speaker Change: So do you see AI and things like al I'm really bringing us back together.

Speaker Change: You mean, the World America Inside America or the world in total it sounds like there's so many fractions.

<unk>.

Speaker Change: I'm going to leave the Americans.

Speaker Change: We are divided which is an issue, but then maybe we get a lot of growth and suddenly we're less divided.

Speaker Change: But I don't think on the international stage, it's going to get better it's going to get much worse.

Unknown Attendee: 0.2 percent is being spent on what you could call, what they're calling, AI. And then software really depends on who builds it. You need money, and you need the right people, and you need the right partners. And it's all three components. And, of course, America, without going into details, is pretty far ahead on certain elements. But our main adversaries are aware of this. And in my view, and again, no one has to agree with me in this room, but I think you get world peace by the U.S. being clearly the strongest nation in the world.

Speaker Change: And then that's a separate discussion for me I'm spending a lot of my time meeting with people in D C.

Speaker Change: <unk> begun pleading you know theres, some 0.2% of the Dod budget, 0.2% now.

Speaker Change: Now I'm, not making that up 0.2% is being spent on.

Speaker Change: What you could call what they're calling AI and then what you know software it really depends on who built it it's not it's you need money and you need the right people and you need the right partners and there's it's all three components and of course America without going into details is.

Pretty far ahead on certain elements, but.

Speaker Change: Our main adversaries are are aware of this and.

Speaker Change: This in my view and again no one has to agree with me in this room, but I think you get world peace by U S being clearly the strongest nation in the world.

Unknown Attendee: Because you could argue about our pluses and minuses and faults. I think we spend a lot of time talking about that. But I'm not aware of any other country that is as benevolent with great power as the U.S. And those of you in this room and others who disagree with me, all right, so I could just narrow it down.

Speaker Change: Because were for.

Speaker Change: You could argue about our pluses and minuses and false I think we spent a lot of time talking about that.

Speaker Change: But.

Speaker Change: I'm not aware of any other country that is a benevolent with great power as the U S.

Speaker Change: And those of you in this room and others, who disagree with it.

Unknown Attendee: Our adversaries will be very different than we would be if they had military superiority in AI. So that's a separate issue, but I think that's how you get a more unified world. But if I had to predict, it's going to be pretty rough. Dr. Earp, how's it going?

Speaker Change: Okay. So I can just narrow it our adversaries.

Speaker Change: Be very different than we would be if they had military superiority in AI. So that that's a separate issue, but I think that's how you get.

Speaker Change: A more unified world, but were that's not going to it's going to be my I. If I had to predict is going to be pretty rough.

Speaker Change: Thank you for these questions really appreciate it. Thank you so much.

Speaker Change: Yes.

Please welcome telco managing partner at DXP, John Boehm.

The Doctor I forgot one.

John Thornhill: Very firm.

Earp: Yeah, oh my no-notes perfect so I've been in and around quite a bit of launching of commercial on Foundry a handful of different organizations First time I saw about six seven years ago One of the big things leadership from Palantir is always saying is this technology is at least five years ahead of our next nearest competitors Do you still feel like you've got that kind of a gap in between where you are today and and other competitors and enterprise stats analytic space It depends, you know, we have a number of products certain aspects of our product It's it the five years ahead is off also euphemism for no one really understands why this works, right? And they haven't started building now.

Oh My no notes perfect.

John Thornhill: I've been in and around quite a bit of launching.

John Thornhill: Commercial on foundry handful of different organizations first time I saw about six seven years ago and one of the big things leadership from Pelletier is always saying is this technologies at least five years ahead of our next nearest competitors do you feel like you've got that kind of a gap in between where you are today and in other competitors.

Speaker Change: And enterprise SaaS analytics space it depends.

Speaker Change: Have.

Speaker Change: A number of products certain aspects of our product.

Speaker Change: Five years ahead is also euphemism for no one really understands why this works right and they haven't started building now I would say in the AI context. The context of time has shifted so like people are doing what you would take 10 years in like two years or a year.

Earp: I would say in the AI context, the context of time is shifted. So, like people are doing what you would take ten years and, like, two years or a year. Certain products that aren't relevant for this audience, like PG, Gaia Nexus peering Apollo, might be like this because no one else is going to build them. Then you have Foundry. I don't think I don't think anyone's going to build Foundry for lots of reasons.

Speaker Change: Certain products that arent relevant for this audience.

Like P G Guy.

Speaker Change: Nexus peering.

Speaker Change: Apollo might be like no one else can build them.

Speaker Change: Then you have foundry I don't think I don't think anyone's.

Earp: It's expensive hard to build It only looks really valuable now so like because you it was valuable before but somehow it wasn't in the zeitgeist that it how valuable it could be Foundry until AI was like it was kind of like a God bless all of you who adopted it and it's I believe it is the best product on the market But it it basically was like we will make your enterprise work better, but you're not going to have any fun and so it It and now it's like we will get your enterprise sizzling because if you have it you can adopt AI in an ethical way and Change your business. So now it's kind of for lack of a better word. It's sexy.

Speaker Change: Going to build foundry for lots of reasons, it's expensive hard to build.

Speaker Change: It only looks really valuable now so like because you it was valuable before but somehow it wasn't in the site case that it will.

Speaker Change: How valuable it could be foundry until AI was like it was kind of like a.

Speaker Change: God Bless all of you who adopted it and it's I believe it is the best product on the market, but it is it basically was like are we will make your enterprise work better, but youre not going to have any fund.

Speaker Change: And so it.

Speaker Change: And now it's like we will get your enterprise sizzling, because if you have it you can adopt AI in an ethical way and change your business. So now it's kind of for lack of a better word is sexy.

Earp: So there may actually be people who build components of Foundry, and you know, I cannot believe that people are not focused on building like branching ontology logic that everyone's going to need. But, but you know, one of the things about being kind of, you know, like from the outside, we just look like we're not following a playbook. So people are like, well, but where's the playbook aspect of ontology branching Where does that fit into my five and six?

Speaker Change: No.

Speaker Change: There may actually be people, who build components of foundry and.

I cannot believe that people are not focused on building like branching ontology.

Speaker Change: Logic.

Everyone's going to need, but but you know one of the things about being kind of.

Speaker Change: Like from the outside.

Speaker Change: We just look we're not following our playbook. So people are like well, but where's the playbook aspect of ontology branching where does that fit into my five six whatever however, they'd show.

Earp: whatever, however, they'd so in each one of these cases, we're de facto very far ahead. I would think the five-year thing might have to be compressed that because AI, large-length models, and copilot things will allow you to work quicker, and people will see the roadmap, and quite frankly, If you're looking at PJ, you know, we have I don't know a large percent of the dressable market, but no one cares because the market is very small, right? Whereas here, the market is infinite.

Speaker Change: In each one of these cases.

Speaker Change: We are de facto very far ahead.

Speaker Change: I think the five year thing, Mike you have to compress that because.

AI large items models Copilots are things will allow you to work quicker and people see the roadmap and quite frankly.

Speaker Change: Youre looking at P. J, we have I don't know a large percent of addressable market, but no one cares because the market is very small right, whereas here the market is infinite it's the only market that matters.

Earp: It's the only market that matters I'm still somewhat optimistic because you We owe a lot to the venture community I think the venture community is going to have a very hard time and that they want they want a model where you all the money isn't building new LLMs and I believe all the value is in managing the LLM in a safe effective way and writing logic to it And that's a subtle difference if if we are right You want to be a company that's off the ground and has something like we have existing client base a reputation something like foundry an ability to implement ontology Understanding of tooling the relay even even the understanding of you would need tools It sounds very simple, right and obviously LLM needs needs tools But even that is like somehow outside the fine financing bailiwick of most venture people and then you have large companies And These companies are the best in the world at what they do, but with notable exceptions They in my business the software business. They they they're just not that many software engineers and then It's like so their model is to acquire companies well, and they're not that many companies to acquire So I'm pretty optimistic about this, but on the inside on the other hand internally we are running very quickly You know we don't want to make the mistake of being ahead and then losing right?

Speaker Change: I'm still somewhat optimistic because.

Speaker Change: We owe a lot to the venture community I think the venture community is going to have a very hard time in that they want they want a model where you all the money is in building new LMS.

Speaker Change: And I believe all the value is in managing the AUM in a safe effective way in writing logic to it and that's a subtle difference. If if we are right you wanted to be a company thats off the ground and has something like we have existing client base, our reputation something like foundry and.

<unk> ability to implement ontology.

Understanding of tooling that even even understanding of you would need tools. It sounds very simple right and obviously L. A need needs tools, but even that has like somehow outside to find financing bailiwick of most venture people and then you have large companies.

Speaker Change: Okay.

These companies are the best in the world at what they do but with notable exceptions.

Speaker Change: In my business to software business day date, Theyre, just not that many software engineers and then.

Speaker Change: It's like so their model is to acquire companies well and they're not that many companies to acquire so I'm pretty optimistic about this but on the anti <unk> on the other hand internally we are running very quickly.

Speaker Change: We don't want to make them a stake of being ahead and then losing right. So you were saying about the compression of the time frame. So I think that one of the things right now <unk> are going accessing human source material humans have created that first level of material and so you can take that 10 years of understanding down to one because they are all the source material is human generated how do.

Earp: So you were saying about the compression of the time frame, so I think that one of the things right now LLMs are going to and accessing human source material. Humans have created that first level of material, and so you can take that ten years of understanding down to one because all the source material is human generated. How do we prevent as we layer more AI generated content on top of, you know, out in the ecosystem of the internet? How do we keep these AI models from getting dumber by then consuming their own output and then using that in the next training set?

We prevent as we layer more AI generated content on top of it out in these even in the ecosystem of the Internet how do we keep these AI models from getting Dumber buy then consuming their own output and then using that in the next training set.

Speaker Change: Well, that's a super interesting question. I would say most people are worried about them getting smarter and then making us irrelevant. And I hope we have the dual challenge of that they get dumber and dumber, and then, of course, the way to solve that problem is you need Foundry. You need logic and even ontology, but uh? So that's that's the that's the answer. There's a real chance that you're right that they get dumber and dumber, but Yeah, and again.

Well, that's a super interesting question I would say most people are worried about them getting smarter and then making us irrelevant.

Speaker Change: And I I hope, we have the dual challenge of that they get dumber and Dumber and then of course the way to solve that problem is you need foundry you need logic can eat in oncology, but.

Speaker Change: So.

Speaker Change: The debt that would be the answer.

Sure.

Speaker Change: There's a real chance that you're right that they get dumber and dumber.

Speaker Change: Right.

Speaker Change: I think you should maybe explain that problem to the venture community. That's my critique. That's what they're selling, and You know it's I look, and you'd be surprised. We should not be running around telling people Look, your thing is going to fail. I do it anyway, but partly because no one listens. And it's like, you know, but that is a legitimate and profound critique of what? Most of the businesses are they're being financed here are going to do all right. Thank you. Thank you very much. I think this may be the time to throw me off the stage, and others.

Speaker Change: Yeah, and again I think you should maybe explain that problem to the venture community. That's my critique, that's what they're selling.

Speaker Change: And they now.

Speaker Change: Look I you'd be surprised we should not be running around telling people look you're thinking it's going to fail.

Speaker Change: I do it anyway, but I, partly because no one listens and it's like but that's that that is a legitimate and profound critique of what most of the businesses that are being financed here are going to do alright. Thank you. Thank you very much.

Speaker Change: I think I may be.

Speaker Change: This is the time to throw me off the stage.

Others.

Speaker Change: Please welcome Adrian Miu from Cleveland Clinic. Hi, thanks for having me. My question has to do with your position in AI and what sort of influence you see for patient care and hospital operations. Well, first of all, it's a great honor. I think you were our first partner in your space. Yeah. I just want to give a shout-out, too, to Jeremy, Drew, and Pargal.

Speaker Change: Oh, Yes, I know I can't Okay. Please welcome from Cleveland.

Speaker Change: Obviously.

Speaker Change: Huge honor.

From Cleveland: Hi, Thanks for having me.

Speaker Change: My question has to do with your position in AI.

And what sort of influence you see for patient care.

Speaker Change: Hospital operations.

Speaker Change: Well first of all I'm.

Speaker Change: Great Honor I think your first partner in your space.

Speaker Change: And I just want to give a shout out to the Jeremy drew embargo.

Speaker Change: They're great. They are. And the team below them, I was in New York over the weekend, and for some weird hour, and they're all there.

Speaker Change: They are great.

Speaker Change: Sure.

Speaker Change: Hmm.

Speaker Change: I mean below them very I was in <unk>.

Speaker Change: In New York and over the weekend and some weird hour and they are all there.

Speaker Change: It's a very special crew. Look, in your space, you have governance issues, ethical issues, and privacy issues. Everybody wants effective care, but nobody wants their personal patient data exposed to anyone.

Speaker Change: Various special crew.

Speaker Change: Look in your space you have governance issues you have ethical issues you have privacy issues.

Speaker Change: Nobody everybody wants effective care nobody wants their personal patient data exposed to anyone.

Speaker Change: You have very technical users. Technical users will not use a product that tells them that they're not engaged with, so they need to be brought along. Quite frankly, you have legal liability issues, so how do you protect health care providers from being randomly sued? And anyone who's got a relative or is involved in health care has a somewhat cynical view of how that happens, or a realistic view. And then the stakes couldn't be higher.

Speaker Change: You have very technical users.

Speaker Change: Technical users do not will not use a product that tells them that they're not engaged with so they need to be brought along.

Speaker Change: Quite frankly of legal liability issues. So how do you protect health care providers from being randomly suite.

Speaker Change: And no.

Speaker Change: Anyone who's got.

Speaker Change: Relative or is involved in health care.

As a somewhat cynical view of how that happens.

Speaker Change: Sure.

Speaker Change: Or realistic view.

Speaker Change: And and then mistakes couldnt be higher and so the way you implement is you show the end user that this is reliable it is augmenting their work, it's making their work safer more effective and you get them because only they can tell their peers that this is working to tell their peers. This work.

Speaker Change: And so the way you implement this is you show the end user that this is reliable, it is augmenting their work, it's making their work safer and more effective. And you get them, because only they can tell their peers that this is working, to tell their peers that this works. And then it's obviously very helpful when you have a mission-driven organization like yours, where people up and down the chain are really engaged in this.

Speaker Change: And then it's obviously very helpful. When you have.

Speaker Change: A mission driven organization like yours, where people.

Speaker Change: Up and down the chain are really engaged.

Speaker Change: In this.

Speaker Change: And then I do think in your industry, as I mentioned before, you guys are under a lot of pressure because of high regulation, a litigious context, and margins that are pressed. And there is no answer for that besides implementing software because you can't change the other variables.

Speaker Change: And <unk> and then I do think in your industry as I mentioned before it's you guys are under a lot of pressure because of.

Speaker Change: Hi regulation.

Speaker Change: Litigious context, and Le and margins that are pressed and there is no answer for that besides implementing software.

Speaker Change: It's just that.

Speaker Change: And we could debate – if we could change the other variables, that would also be an option. And quite frankly, if I were in charge of that, I would be very open to discussion. But in reality, those are the constraints for now and probably for a long time.

Speaker Change: Can't change the other variables and we could debate.

Speaker Change: If we could change the other variables that would also be an option and quite frankly, if I were.

Speaker Change: In charge of that I would be very open to discussions but in reality those are the constraints for now and probably for a long time and then you have to get per verbally a lot more from less.

Speaker Change: And then you have to get proverbially a lot more from less. And that's what Foundry does. That's what AIP does. And that's why our partnerships in this area have gone very well. Yeah. Can I just add to that?

Speaker Change: And that is what foundry does thats, where AIP does.

Speaker Change: And that's why our partnerships in this area have gone very well, yeah can I just add to that.

Speaker Change: Sort of my philosophy with what I do is to try to be disruptive, think outside the box. Basically, whatever I see what's being currently done, I want to do the opposite. And so I'm trying to leverage you guys and AI and the technologies. In your position, what would you say is the next step towards operations? How to leverage that AI technology in the hospital industry?

Speaker Change: The sort of my philosophy with what I do is to try to be disruptive.

Speaker Change: Think outside the box.

Speaker Change: Basically whatever I see what's been currently done I want it to the opposite.

Speaker Change: And so I'm trying to leverage you guys in AI and the technologies.

Speaker Change: In your position.

Speaker Change: What would you say is the next step towards operations, how to leverage that technology in house the hospital industry.

Speaker Change: Well, I mean, it's obvious in some ways, like identifying people's risk equivities, having an updated understanding of the literature and the treatment profiles, being able to give the end user the absolute most accurate current understanding of patient treatment data. We are probably not very far from more individualized or already individualized AI and broadly defined hospitals without AI and broadly defined. And I would say, you know, in a governance structure that protects everyone involved. Because, you know, you can't do individualized medicine without making an individualized assessment.

Speaker Change: Well I mean, it's obvious and like in some ways like identify people's risk liquidities, having an updated understanding of the ore of the literature and the treatment profiles being able to give the end user the absolute most accurate understand current understanding of patient treatment data.

Speaker Change: Probably not very far from more individualized or already individualized.

Speaker Change: AI and.

Speaker Change: Broadly if a hospital without AI broadly.

Speaker Change: Broadly defined.

Speaker Change: And I would say.

Speaker Change: And a governance structure that protects everyone involved.

Speaker Change: Because you can't do individualized medicine, without making it an individualized assessment Matt.

Speaker Change: And that without the right governance structure puts the health care provider in a misaligned relationship with the person they're trying to treat. And there's a finality to health care, but it's like you have to have full alignment between patient and doctor. But then you're doing this in a data-rich, regulated environment. It's a very hard use case. So, thank you for your time. Thank you. Thank you. Thank you. And welcome. Welcome to our humble abode.

Speaker Change: Without the right governance structure puts the healthcare provider in our mis aligned relationship with the person, they're trying to treat and does the banality of health care, but it's like you have to have full alignment between patient and doctor.

Speaker Change: But then youre doing this in a data rich regulated environment.

Speaker Change: So it's a very hard use case.

Speaker Change: Thank you for your time thank you.

Speaker Change: Yes.

Speaker Change: Sure.

Speaker Change: And welcome welcome to our humble Abboud and I Hope you learn a lot we will learn a lot from you.

Speaker Change: And thanks for coming.

Speaker Change: Okay.

Speaker Change: And I hope you learn a lot. We will learn a lot from you, and thanks for coming. Please welcome from Palantir, Chief Technology Officer Sham Sankar. Good morning.

Speaker Change: Please welcome from Palin Tier Chief Technology Officer, Sean Sungard.

Speaker Change: Yes.

Speaker Change: Yes.

Speaker Change: Okay.

Speaker Change: Okay.

Sean Sungard: Good morning.

Shyam Sankar: All right, so June of this year, you know, from this stage, I launched AIP, so just, I don't know, three, four months ago here, which AIP is our open, extensible AI platform to help you build AI-enabled applications by bringing LLMs to your private data on your private networks that power these experiences both safely but also efficiently. But really, the observation that we've had over this period of time is that the battleground of AI is the factory floor, it's the front line. And that's because AI transformation is an empirical journey. You have to experience it for yourself. You can't think your way through it. You have to go and do it.

Sean Sungard: Alright so.

Sean Sungard: June of this year from this stage I launched AIP, there's just three or four months ago here.

Sean Sungard: Which is our open extensible AI platform to help you build AI enabled applications by bringing llm's to your private data on your private networks.

Sean Sungard: That power these experiences, but safely but also efficacious Lee.

But really the observation that we've had over this period of time is that the battleground of AI. It's the factory floor, it's the frontline and Thats because AI transformation is an empirical journey.

Sean Sungard: You have to experience. It you can't think your way through it you have to go do it.

Shyam Sankar: And so today, we have time. We have tons of product to show, but we're going to have our customers actually show all the new things that we've done together. We're going to have them take you on their experiential journeys here. Customers like HCA, who have proactively looked into the future of their actual schedules to identify patient bottlenecks and staffing shortages and solve them upstream. Or Eaton, where they've gone from an alert inbox that tells them things that are going wrong to a solution inbox that tells them things that they can be doing about things that have not yet gone wrong but might. The important thing about all of these customers is that they have pushed to prod. I tell every Palatarian internally, like, I don't want to see proof of concepts. I want to see evidence.

Sean Sungard: And so today, we have tons of product to show, but we're going to have our customers actually show all of the new things that we've done together and we're in it have them take you on their experiential journeys here customers like HCA, who have proactively looked into the future of their actual schedules to identify patient bottlenecks and staffing shortages and solve them up.

Sean Sungard: Stream or Eaton, where they've gone from an alert inbox that tells them things that are going wrong to a solution and box. It tells them things that they can be doing about things that have not yet gone wrong, but might.

Sean Sungard: The important thing about all of these customers is that they have pushed to prod.

Speaker Change: I know every pound here internally like I don't want to see proof of concept I want to see proof.

Shyam Sankar: And that's largely because these LLMs are different. They are something new. They break our existing mental models, and I thought that's what I'd really focus on sharing with you. What are the ways in which we think about these things, and how are they kind of different from what's come before? And we could start there by asking, what is an LLM doing? Or perhaps more importantly, what is it not doing? Well, it's not doing algorithmic reasoning, which is a form of computation that's so precise that there's no ambiguity in its execution. Right?

Speaker Change: And that's largely because these <unk> are different they are something new they break our existing mental models and I thought that's what I'd really focus on sharing with you what are the ways in which we think about these things and how are they kind of different than what's come before.

Speaker Change: And we could start there by asking what does an ela I'm doing or perhaps more importantly, what does it not doing.

Speaker Change: It's not doing algorithmic reasoning, which is a form of computation. That's so precise that theres no ambiguity in its execution and I like to the extent it doesn't do what you said that we call. It a bug it that's not supposed to happen now.

Shyam Sankar: Like, to the extent that it doesn't do what you said, we call that a bug. That's not supposed to happen. And on the other extreme, it's not doing human thought. It is an inherently creative medium that is wrought with ambiguity. That's part of the feature of human thought. It occupies this middle ground in between these two things. And it can be very tricky as you're implementing these things to hold that tension in a precise way because it is fluent in natural language, but it doesn't actually understand what it's saying. So it's not human thought.

Speaker Change: And on the other extreme it's not doing human thought.

Speaker Change: Eight eight inherently creative medium that is wrought with ambiguity that's part of the feature of human thought it occupies this middle ground in between these two things here and it can be very tricky as you're implementing these things to hold that tension in a precise way because.

Speaker Change: It is fluid and natural language, but it doesn't actually understand what I'm, saying, so it's not human thought.

Speaker Change: And while it is instructable, an ordinary human pros, it's horrible at algorithmic reasoning.

Shyam Sankar: And while it is instructable in ordinary human prose, it's horrible at algorithmic reasoning. So it's this third type of thing. And I think one of the characteristics of this third type of thing is that LLMs are statistics, not calculus. It is this kind of stochastic. genius, if you will. There's a lot of randomness in it. The best metaphor for this, and perhaps one of the things to think about as you're implementing these use cases, is what it was like for us when we first started trying to predict the weather. When we first started trying to predict the weather, I think, in the rough mid-1800s, we thought it was going to be like predicting the next eclipse.

Speaker Change: So it's this third type of thing and I think one of the characteristics of this third type of thing is that <unk> are statistics not calculus. It is this kind of stochastic Jean if you will if there is a lot of randomness in it the best metaphor for this and perhaps one of the things to think about as you're implementing these use cases.

Speaker Change: Is what it was like for US when we first started trying to predict the weather. When we first started trying to predict the weather I think in the rough mid 18, hundreds we thought it was going to be like predicting the next eclipse like once we figure out the math behind this thing I'm going to tell you precisely what the weather's going to be in this location 100 years from now at this time and obviously intuitively, we all understand that's not how weather works.

Shyam Sankar: Like once we figure out the math behind this thing, I'm going to tell you precisely what the weather is going to be like in this location 100 years from now at this time. And, obviously, intuitively, we all understand that's not how weather works, right?

Speaker Change: Alright, well astronomy is calculus, it's governed by the truth of the underlying mathematical equation I can tell you that in the year 2086.

Shyam Sankar: Astronomy is calculus. It's governed by the truth of the underlying mathematical equation. For example, I can tell you that in the year 2186, I can tell you exactly where you need to be on Earth to see the longest eclipse in the last 10,000 years. That's the nature of calculus. On the other hand, with weather, it's more like it might rain today.

Speaker Change: I can tell you exactly where you need to be on Earth to see the longest eclipsing the left last 10000 years.

Speaker Change: And that's the nature of calculus.

Speaker Change: And with whether it's more like it might rain today alright.

Shyam Sankar: The equations are dominated by the propagation of era and some underlying fundamental stochasticity, or randomness in it. And we see this through another lens, too, like the fixation with pushing to prod comes by thinking about this by analogy. When you introduce even one stochastic variable into what was otherwise a deterministic system, your software, you've now made the entire system stochastic, and that requires us to rethink some fundamental things that are in the tool chain, how we as humans reason about the problem, and how we incorporate it to achieve outcomes. This is the difference between a demo and a product, right? So we had a self-driving car demo in 2025. They drove for 132 miles.

Speaker Change: The equations are dominated by the propagation of error and some underlying fundamentals stochastic city randomness in it.

Speaker Change: Okay.

Speaker Change: And we see.

Speaker Change: This through another lens to like be fixation with pushing to prod comes by thinking about this by analogy when you introduce even one stochastic variable into what was otherwise a deterministic system your software.

You've now made the entire system stochastic and that requires us to rethink some fundamental things that are in the tool chain, how we as human reason about the problem and how we incorporate it to achieve outcomes.

Speaker Change: This is the difference between a demo and a product right. So we had a self driving car demo in 2025 to 232 months like it it's not around the block trades at a meaningful distance here and arguably you could say maybe this year, we will have a product maybe maybe it will take another year or two more net after a 44 million self driving my.

Shyam Sankar: It's not around the block. It's a meaningful distance here. Arguably, you could say maybe this year will have a product. Maybe. Maybe it'll take another year or two, more. That's after 44 million self-driving miles.

Shyam Sankar: And some of that need to have all that experience speaks to this intuition that we have, that there's something so castic about it. We need a lot of proof to understand how it's going to behave. So, you know, in some sense, it's kind of easy to build a demo. It's much harder to build a product, and that's what we're focused on. And when you really think about that, you start to uncover the fundamental primitives and infrastructure that you're going to need to harness the true underlying value of transformation that LLMs can bring to you here, and realize that all of that value comes from the elegant integration of algorithmic reasoning, the LLM, and human thought. And that's what you're going to see.

Speaker Change: And some of that they need to have all that experienced speaks to this intuition that we have that theres something so caustic about it we need a lot of proof to understand how it's going to behave.

Speaker Change: So sometimes it's kind of easy to build a demo it's much harder to build the product and that's what we're focused on and when you really think about that you start to uncover the fundamental primitives in infrastructure that you're going to need to harness the true underlying value of transformation that <unk> can bring to you here and realizing that all of that value.

Speaker Change: It's from the elegant integration of the algorithmic reasoning, the LLM and human thought on that.

Speaker Change: What youre going to see in the workflows to customers.

Speaker Change: So off later today.

Shyam Sankar: in the workflows that customers will show off later today. But one of the core ways to leverage the power, the underlying power of the LLM, is this idea of the K-LLM kernel. Why would you use one LLM when you can use K? We can kind of think about, you know, if you squint and just go with this analogy for a second, like each of these LLMs is like a slightly deranged mad scientist or mad genius. You know, when we have this sort of situation as humans, we build committees.

Speaker Change: But one of the core ways to leverage the power of the underlying power of the <unk> is this idea of the KLM carnell.

Speaker Change: Why would you use one LLM when you can use K.

We can kind of think about if you squint and just go with this analogy for a second like each of these <unk> is like a slightly deranged mad scientist Mad genius.

Speaker Change: When we have this sort of situation as humans, we build committees why would you ask one mad genius. The question. When you ask a committee of Mad Genius is the question and then synthesize the answer that youre going to get back and understand the kind of a rich context, where they agree where they disagree and what you want to do about that.

Shyam Sankar: Why would you ask one mad genius the question when you can ask a committee of mad geniuses the question and then synthesize the answer that you're going to get back and understand the kind of rich context where they agree, where they disagree, and what you want to do about that? It also forces you to realize the fundamental truth here that there isn't an answer in a context like this. There are answers.

Speaker Change: It also forces you to realize the fundamental truth here that there isn't an answer in a context like this there are answers.

And it's probably most important for the underlying application developers to realize that because if you if youre in this simple paradigm pretending that this is deterministic asking a question getting back a single answer that could be a hallucination that could be kind of it's not the answer it's an answer and then youre surfacing that to the user you're going to find yourself stuck somewhere.

Shyam Sankar: And it's probably most important for the underlying application developers to realize that. Because if you're in this simple paradigm, pretending that this is deterministic, asking a question, getting back a single answer that could be a hallucination, that could be kind of, it's not the answer, it's an answer, and then you're surfacing that to the user, you're going to find yourself stuck somewhere on that journey between 2005 and 2023.

Shyam Sankar: If instead you're using a kernel, like the KLM kernel, and you're getting back multiple answers, you're synthesizing those answers, you're able to flow that through and drive the entire user experience towards operational outcomes. It's how you synthesize, analyze, and design the software that's really going to matter here. So if we just walk through this approach briefly so that we understand what we're looking at, you would take your prompt, and you would feed it to the K-LMs. In this case, I just have four to keep it simple and reason through it.

Speaker Change: We are in that journey between 2005 and 2023, if instead you are using a kernel like the KLM kernel and Youre getting back multiple answers you're synthesizing those answers youre able to flow that through and drive the entire user experience towards operational outcomes.

Speaker Change: How you synthesize analyze and designed the software that's really going to matter here.

Speaker Change: So if we just walked through this approach briefly so that we understand we're looking at you would take your prompt and you would feed it to K L. Happens in this case I just have four to keep it simple and in reading through it you get back the responses from each of these LMS you can use that to summarize the response of each LLM and you send the entirety of each of the responses to a synthesis.

Shyam Sankar: You get back responses from each of these LLMs. You can use that to summarize the response of each LLM, and you send the entirety of each of the responses to a synthesis stage to understand, rate, compare the answer that you've gotten and come up with the best possible answer, the dissenting view of the answers, which models got what right, which models got what wrong, and then surface that to both the developer to drive the user experience and the end user as a consequence of doing that.

Speaker Change: <unk> to understand rate compare the answers that you've gotten in come forth with a best possible answer the dissenting viewed the answers which models got what right, which models got what wrong and then surface that you both the developer to drive the user experience and the end user as a consequence of doing that.

Shyam Sankar: Now, the synthesis can be deterministic. If you're getting back JSON or code, you could treat these as abstract syntax trees, you can do formal disks of what you're getting back, and it's a deterministic sense of what's the same and different.

Speaker Change: The synthesis can be deterministic if youre getting back Jay saw on our code you can treat these abstracts in tax rates you can do formal deaths of what youre getting back as the deterministic sense of what's the same and different.

Shyam Sankar: But also, the synthesis can be an LLM step itself, where you have a synthesis prompt that's helping you in the context of the workflow you're trying to do, create the right synthesis that matters for the work that you're doing in that stage. I'm going to illustrate this in a second, but I think what you will see is that the KLLM approach is the only way to manage model stability, operational risk, and harness the true nature of the LLM to win. So when we talk about model stability, if you remember a few months ago, there were reports that CHAPT or GPT4 suddenly got worse or better or different from certain things.

Speaker Change: But also the synthesis can be an L. M step itself, where you have a synthesis prompt that's helping you in.

Speaker Change: In the context of the workflow youre trying to do create the right synthesis that matters for the work that youre doing in that stage.

Illustrate this in a second but I think what youll see is that the KLM approach is the only way to manage model stability operational risk and harness the true nature of the LLM to win.

Speaker Change: So when we talk about model stability. If you remember a few months ago. There were reports that chat TBD or GBT for suddenly got worse or better or different that certain things and if you ask open AI. They would say no nothing changed I always told you these models where stochastic.

Shyam Sankar: And if you asked Open AI, they would say, no, nothing changed. I always told you these models were stochastic. So yeah, you're getting different answers.

Speaker Change: So, yes, youre getting different answers that's the nature of the model I think that's a pretty strong argument for why you can't have one LLM, making critical decisions you need K on the other hand, if you think Oh actually something did change in the underlying model.

Shyam Sankar: That's the nature of the model. I think that's a pretty strong argument for why you can't have one LLM making critical decisions. You need K. On the other hand, if you think, oh, actually, something did change in the underlying model, I think it's a pretty strong argument for why you need more than one LLM; you need K, if you're making a critical decision.

Speaker Change: It's a pretty strong argument for why you need more than one LLS you need K, if you're making a critical decision. So this allows you to triangulate. The answers you are getting back and had a scaffolding around the critical decisions you are trying to drive in your enterprise.

Shyam Sankar: So this allows you to triangulate the answers you're getting back and have a scaffolding around the critical decisions you're trying to drive in your enterprise operationally. Then, if we think about it, there's a lot of interest in creating custom fine-tuned models. That's great.

Speaker Change: Operationally.

Speaker Change: Then if we think about there's a lot of interest in creating custom fine tune models. That's great I think we're going to live in a world, where there's a menagerie of models in the future.

Shyam Sankar: I think we're going to live in a world where there are a menagerie of models in the future. How are you going to deploy them? You know, you think about how you are going to have to do the equivalent of 44 million self-driving miles of testing and evaluation to understand and characterize putting this model and to something as critical as manufacturing batteries or critical national infrastructure. I don't think that's a viable surface area of solution.

Speaker Change: How are you going to deploy them.

Speaker Change: You think about how are you going to have to do the equivalent of 44 million self driving miles of testing and evaluation to understand and characterize putting this model into something is as critical as manufacturing batteries or critical national infrastructure, I don't think thats, a viable surface area of solution.

Shyam Sankar: Why don't you just make it the K-plus-1 model? It's the K-plus-1 LLM that's being evaluated in parallel. You're able to triangulate how it's the same or different in how it performs against a broad surface area. And certainly, contextually, against the use cases you actually built this custom fine-tuned model to perform against. So it's the fastest way. And also, the safe

Speaker Change: I don't you just make it the K plus one month. It is the case with one L. M. That's being evaluated in parallel youre able to triangulate, how it's the same or different in how it performs against a broad surface area and certainly contextually against the use cases, you actually built this custom fine tune model to perform against so it's the fastest way.

Shyam Sankar: This way to get operational value from the models that you want to build in the menagerie here. And then, of course, a big part of this approach is recognizing that you don't have many priors, or you have no priors on these models. How much experience do you have with this committee of experts?

Speaker Change: And also the safest way to get operational value from the models that you want to build in the menagerie here.

And then of course a big.

Speaker Change: Part of this approach is recognizing that you don't have many prior so you have no priors on these models how much experience do you have with this committee of experts overtime, just like a human committee of experts as you get to know the experts better you will have a stronger opinion in this case backed up by the telemetry within AIP of historical responses of which model is going to be.

Shyam Sankar: Over time, just like with a human committee of experts, as you get to know the experts better, you will have a stronger opinion, in this case backed up by the telemetry within AIP of historical responses, of which model is going to be best for what sort of use case. So are you going to ask every question to K models all the time? No.

Speaker Change: Best at what sort of use case. So are you going to ask every question Dk models, all the time now.

Shyam Sankar: You know, as you have stronger and stronger conviction of where to go for what, you'll tune that down. You can tune it down for cost, You can tune it down for correctness and accuracy, but all those levers will be within the control of the enterprise here. So maybe to walk through an example here, the first one shows us using GPT4, Star Chat, Beta, and Lama, to 13B. What models don't really matter per se, but just so you have some context here. So the underlying prompt is that I would like to get back a Python function that takes in this table of employees with columns like first name, last name, year of employment, and company.

Speaker Change: As you have stronger and stronger conviction of where to go for what Youll tune that down you can tune it down for cost you can tune it down for correctness and accuracy, but all of those levers will be within the control of the enterprise here.

Shyam Sankar: And it gives me back the employee at each company with the highest tenure. So the prompt is being sent off to all three models at once. You see the GPT4 answer streaming back in.

So maybe to walk through an example here.

Speaker Change: The first.

Speaker Change: So this shows us using GPT four star Cat Beta and Lama to 13 B.

Speaker Change: What models don't really matter per se, but just to give some context here. So the underlying prompt is that I would like to get back a python function that takes in this table of employees with call ups like first name last name your employment data <unk> company and it gives me back the employee at each company with the highest tenure.

Speaker Change: So the prompt is being sent off to all three models that once you see the GPT for answer screaming back in the other and once you get the full answer you'll get a summary of the answers so it collapses down.

Shyam Sankar: The other, and once you get the full answer, you'll get a summary of the answer, so it collapses down. You see the Star Chat summary, and Lama will be coming shortly.

Speaker Change: You see the starts at summary, and Lama will be coming shortly once all all of these answers come.

Shyam Sankar: Once all these answers come in, they will be sent to the synthesis stage, then to compare and contrast them. And that's where the real magic of the KLLM kernel comes in. So here is the synthesized answer streaming back in here. I'm getting the model's assessment of what code should I actually consider, what Python code should I actually do that meets these requirements? But, much more interestingly, now we see the relative performance of the models. So how did Lama 13 do?

Speaker Change: They will be sent to the synthesis stage, then to compare and contrast them and that's where the real magic of the KLM Carnal comes in.

Speaker Change: So here's the synthesis answer streaming back in here and I'm getting the models assessment of what code should I actually consider what Python code should I actually do that meets these requirements here, but much more interestingly now we see the relative performance of the models. So how did Lama 13, do how to start that do and we see that in the kind of upper right.

Shyam Sankar: How do we do Star Chat? And we see that in the kind of upper right and kind of in the lower middle section here. So let's step through each of these in a second. Here are the summary answers. Now, here's the output from the synthesis. Well, what we can see is that, interestingly, Lama did something where it dropped all employees who had a tenure of less than a year. That was not an explicit requirement of the number.

Speaker Change: And it kind of in the middle and the lower Middle section here.

Speaker Change: So let's step through each of these in a second.

Speaker Change: Here are the summary answers.

Speaker Change: Now here's the output from the synthesis or what we can see is that interestingly Lama did something where it dropped all employees, who had a tenure of less than a year that was not an explicit requirement of the prompt it's kind of interesting at the developer maybe I want to consider that it also shows you the nature of of the stochastic city, that's underlying that so either meet Myra.

Shyam Sankar: the prompt. It's kind of interesting, though, as a developer, maybe I want to consider that. It also shows you the nature of the socasticity that's underlying this. So it either meets my requirements, it considers an edge case I hadn't thought of, or it doesn't. Star Chat actually failed to account for ranking the tenure of the employees.

Speaker Change: <unk>. It considers an edge case I had thought of or it doesn't start that actually failed to account for ranking the tenure of the employees and so this gives me a sense again of this as another dot and understanding how this models when there for them.

Shyam Sankar: And so this gives me a sense, again, of this is another dot in understanding how this model is going to form over the fullness of time here. But I, as the developer or the end user here, have a rich understanding of what I should be thinking about and a much richer understanding than if I just asked one of these models a question. Even if I got quote unquote the right answer, I have a much more varied topology, a 3D understanding of what's going on. And then operationally, one of the interesting things is, well, what if I only had one LLM? Even then, this is still valuable.

Speaker Change: Over the phone or some time here, but I as the as the developer or the end user here have a rich understanding of what I should be thinking about in a much richer understanding that if I just ask one of these models. The question, even if I got quote unquote. The right answer I have a much more very topology of three D understanding of what's going on.

Speaker Change: And then operationally one of the interesting things as well what if I only have one LLM. Even then this is still valuable to think about it is K by one llm's, where we've put this in production with manufacturing customers, where they're trying to leverage their internal enterprise knowledge repositories and create abilities for it.

Shyam Sankar: Think about it as K-by-1 LLMs, where we put this in production with manufacturing customers, where they're trying to leverage their internal enterprise knowledge repositories and create abilities for... frontline workers to interact with it. You can actually take your question and send it to, in parallel, the same LLM at scale and understand how varied, how much concurrency is there in the responses that you're getting back, how much divergence is there. I have a silly example here where I'm just saying, look, pick, basically, I have three doors; which one should I pick?

Speaker Change: Frontline workers to interact with it.

Speaker Change: And you can actually take your question and send it to in parallel the same LLM.

Speaker Change: At scale and understand how varied how much concurrent is there and the responses that youre getting back how much divergence is there.

Speaker Change: I have a silly example, here where I'm, just saying, let's pick basically I have three doors, which one should I pick and.

Shyam Sankar: And this is a real example where I get back a different answer every time I ask. You could kind of maybe expect that in the context of having the stochastic genius. This is your answer. One thing it tells you is this is probably not a good question to ask an LLM. This is an LLM hard problem.

Speaker Change: In the real.

Speaker Change: Example, where I get back a different answer every time I ask you could kind of maybe expect that in the context of having the stochastic genius you answer one thing. It tells you is this is probably not a good question to ask and Ella. This is an LLM hard problem, if you're if you're never going to get convergence on the answers you're asking.

Shyam Sankar: If you're never going to get convergence on the answers you're asking, you know, maybe you're kind of designed this wrong. Some way of thinking about that. But importantly, then, this surfaces, how do I bring this into the workflow? Do I need more knowledge augmentation on the front end? How do I, what sort of algorithmic tools can I build to bring to bear? How do I bring this into production?

Speaker Change: Maybe you kind of design this wrong in some way of thinking about that but importantly that this surfaces, how do I bring this to the workflow do I need more knowledge augmentation on the front end, how do I, what sort of asthmatic tools can I build to bring to bear how do I bring this into production and this has been very valuable where it deeply cuts down hallucinations and there.

Shyam Sankar: And this has been very valuable where it deeply cuts down hallucinations and the reliability of the underlying use cases. A lot of the work I just showed here focuses on chat. But I think you've had the unfortunate experience of having me sit with you and talk about these LLMs for a bit, because you know one thing, I'm highly biased against chat. I think it is the most limiting interface, and it kind of structurally means that we're not going to fully harness the value of LLMs because it curtails our human creativity on what these things can do. The reason that's the key. In this case,

Speaker Change: Reliability of the underlying use cases.

Speaker Change: A lot of the work I just showed here focuses on chat, but if you had the unfortunate experience of having me sit at you and talk about these elements were a bit as you know one thing is I'm highly biased against chat I think it is the most limiting interface and it kind of structurally it means that we're not going to fully harnessed.

Speaker Change: The value of Llm's, because they take curtails, our human creativity on what these things can do.

Shyam Sankar: It's the number one mistake of thinking about, you know, what are you relying on the parametric knowledge of the model to do? The parametric knowledge is like the knowledge that's implicit in the model. What I think is special about these models is that they speak a language, a regular grammar. Yeah, sure, English, but much more interestingly, they speak domain-specific languages.

Speaker Change: The reason that's the case. It is this number one mistake about thinking about what are you relying on the parametric knowledge of the model to do the parametric knowledge is like the knowledge that's implicit in the model.

Speaker Change: What I think is special about these models is that they speak.

Speaker Change: Our language a regular grammar, yes, sure English, but much more interestingly they speak domain specific languages. They speak coat they speak Jason if I could just say it colloquially they speak code and clicks and they do that really really well.

Shyam Sankar: They speak code. They speak JSON. If I could just say it colloquially, they speak code and clicks. And they do that really, really well. On the other hand, if you're trying to somehow jam the specific knowledge of your domain into these models, I think you will find that that's a Sisyphian task. You will find that you never fully get there.

Speaker Change: On the other hand, if you're trying to somehow jam the specific knowledge of your domain into these models I think you will find that the Sisyphean task you will find that you never fully get there they never get better than your true experts or you're always chasing your tail and it's not like a database where you can just jam facts in there. So there it's kind of going back to this calculus.

Shyam Sankar: They never get better than your true experts, or you're always chasing your tail. And it's not like a database where you can just jam facts in there. So it's kind of going back to this calculus statistics thing, that it's like the pressure in the balloon that's kind of moving around. You're going to be facing tradeoffs on the frontier. But here, they are phenomenal.

Speaker Change: Statistics thing that's like the pressure in the balloon is kind of moving around.

Speaker Change: Going to be facing tradeoffs on the frontier.

Speaker Change: But here they are phenomenal. So if you if you thought about this.

Shyam Sankar: So if you think about this as LLM should be defining the experience your humans have when they're using their enterprise tools, then you start to see different patterns emerge here. As this video illustrates, we have this application in the upper right-hand corner here that is our operational application. That application has some sort of internal representation, some sort of JSON that represents the app, and that's what you're going to see kind of rendered here. Okay, so it looks like computer speak. Well, I would like to add properties to this table that are every location property that's in the ontology.

Speaker Change: As LLM should be defining the experience you're humans have when they're using their enterprise tools.

Speaker Change: Then you start to see different different patterns emerge here so.

Shyam Sankar: Now, I can do that as a human. It's probably on the order of 50 to 100 clicks to go into the application, edit it, and add all of these things. Or I can just say this, and actually, the LLM can speak those clicks, if you will, and change my underlying application state for me. This is, I think, a better way of thinking about it. the paradigms that we could be using to power these sorts of experiences.

Speaker Change: This video illustrate we have this application in the upper right hand corner here that is our operational application.

Speaker Change: That application has some sort of internal representation, so I'm sort of Jason that represents the app and that's what you're going to see kind of rendered here. Okay. So it looks like computer speak.

Speaker Change: Well I would like to add properties to this table that our every location property that's in the oncology.

Speaker Change: Now I can do that as a human it's probably on the order of 50 to 100 clicks to go into the application to edit to add all of these things where I can just say this and actually the L. M can speak those clicks, if you will and change my underlying application state for me.

Speaker Change: This is I think a better way of thinking about the paradigm that we could be using to power. These sorts of experience is a concrete manifestation of how the experience. The L. M creates changes the human interaction with software.

Shyam Sankar: It's a concrete manifestation of how the experience the LLM creates changes the human interaction with software and allows us to leverage tools. So LLMs need software tools. They need that augmentation to do algorithmic reasoning. That is how you get the precision of calculus with the power of statistics.

Speaker Change: And allows us to leverage tools, so llm's need software tools, they need that augmentation to do outback reasoning that is how you get the precision of calculus with the power of statistics.

Shyam Sankar: And all of the value sits at that intersection of your deterministic code, the stochastic LLM. This is how you register your existing AI models, your Jupiter notebooks, your Lambda functions, your high-performance compute clusters. How do I take all of that deterministic value I have and unleash it in this sort of paradigm here? And I think, in many ways, chat also limits us to thinking about LLMs as a single tool. It's like hitting in music; it's like hitting one note, right?

Speaker Change: And all of the value sits at that intersection of door deterministic code and the stochastic LLS.

Speaker Change: This is how you register your existing AI models your Jupiter notebooks. Your Lambda functions. Your high performance compute clusters, how do I take all of that deterministic value I have and unleash it in this sort of paradigm here and I think in many ways chat also limits us to thinking about <unk> as a single tool it's like hitting in.

Speaker Change: Music, it's like hitting one note right like you want to be playing a court and the court is the elegant integration of human thought algo.

Shyam Sankar: Like you want to be playing a chord. And that chord is the elegant integration of human thought algorithmic reasoning, these tools that you have, and the LLMs. And so I want to just signpost this to really watch for this in the customer demonstrations later. And here's an example that we'll see from Eaton where we can look at, what do these tools mean? What do they look like?

Speaker Change: Algemene reasoning these tools that you have and yellow labs.

Speaker Change: And so I want to just sign posts this fit to really watch for this in the customer demonstrations later and here's an example that we will see from Eaton, where we can look at what are these tools I mean, what do they look like so we start off when we see the prompt here. This is of course, the problem thats going to an LLS, but underneath the profit you can see it's been given a tool the ability to query the.

Shyam Sankar: So we start off, and this is, of course, the prompt that's going to an LLM. But underneath the prompt, you can see it's been given a tool, the ability to query the ontology for plant material, I think. You see more. And then the interplay, okay, well, I can query for plants and plant material as well.

Speaker Change: G four plant material I think.

Speaker Change: You see more prompts and then the interplay, okay, well I can query for plants and four.

Speaker Change: Plant material as well.

Shyam Sankar: And as we kind of go on here, this is like chained logic. This is an LLM-backed function. You know, I'm building logic that's running my enterprise. It's fluidly going between the LLM and a query tool. And now what we see, it's also going to an ontology function. You can know that that's the function that's calling the AI model. It's calling the Lambda function. It's calling the notebook.

Speaker Change: And as we kind of go on here. This is like change. The logic. This is an elegant backed function I'm building logic, that's running my enterprise it fluidly going between the L M and a query tool and now what we see it's also going to an ontology function. That's the function that's calling the AI model, it's calling the lambda function, it's calling the notebook.

Shyam Sankar: It's calling the logic in my enterprise, and it's letting the LLM wield it in an operational context here with the gargrails, with the safety, in an elegant sort of way. So I think you'll see that in each of these demonstrations. How did we build, how did the customers build these functions, and how are they working between the LLMs and the algorithmic tooling? And that's a big part of this. What will also come through in these demonstrations is that the LLMs don't change what problems matter to your business.

Speaker Change: The logic in my enterprise and it's letting the LLM wheeled it in an operational context here with the guard rails with the safety in an elegant sort of way so.

Speaker Change: You'll see that each of these demonstrations how did we build how do the customers to build these functions and how are they going between the <unk> and the augment tooling.

Speaker Change: And that's a big part of this what will also come through in these demonstrations of the <unk> don't change what problems matter to your business the promise that matter your business or the same problems that matter to your business yesterday. They allow you to solve these problems substantially more quickly.

Shyam Sankar: The problems that matter to your business are the same problems that mattered to your business yesterday. They allow you to solve these problems substantially more quickly. And I think that's really exciting. Because with AI, as Alex even kind of alluded to, all the value really accretes to the incumbent, the incumbent users, the incumbent applications, the incumbent organizations.

Speaker Change: And I think that that's really except because with AI as Alex even kind of alluded to all the value really accrete to the incumbent the incumbent users the incumbent applications the incumbent organizations.

Shyam Sankar: And that means you can just solve these things substantially more quickly. Okay, so I think we've had enough of our mental models. Let's get to the front lines. Ted will be up next to tell us and take us there. Like what works? How do you approach it? How do you get started? Thank you, guys. Please welcome Ted Maybrey, head of global commercial, Palantir.

Speaker Change: And that means you can just solve these things substantially more quickly.

Speaker Change: Okay. So I think enough with our mental models, let's get to the Frontlines Ted will be up next to tell us and take US there like what works how do you approach. It how do you get started thank.

Ted: Thank you guys.

Ted: Yes.

Ted: Yes.

Ted: Okay.

Ted: Please welcome from talents here head of global commercial Ted maybe.

Ted: [music].

Ted: Sure.

Ted:

Hi, everybody.