Skip the Megacaps: These 2 Under-the-Radar AI Stocks Have More Room to 10X

Innodata reported record Q2 revenue of $92.1 million, up 58% year over year, while adjusted earnings increased 92%, as it begins shifting from data-labeling services toward a self-serve software platform. BigBear.ai secured a 10-year U.S. Air Force IDIQ contract with a ceiling of up to $900 million, although the award does not guarantee revenue and the company remains deeply unprofitable. Both stocks trade substantially below prior highs—Innodata more than 45% below its roughly $125 peak and BigBear.ai near its $2.60 low versus more than $9 high—leaving upside contingent on platform adoption, contract conversion, and improved profitability.

Analysis

INOD’s rerating hinge is not another services-growth print but proof that platform revenue is recurring, low-touch, and non-cannibalistic. A hyperscaler distribution agreement could rapidly expand reach, but it would likely exchange gross margin for volume and bargaining power; the key disclosure is net revenue retention, platform ACV, and gross-margin progression rather than the number of evaluations. Customer concentration remains the dominant downside transmission mechanism: a single account pause would simultaneously impair revenue growth, utilization, and the market’s willingness to underwrite a software multiple.

BBAI’s contract vehicle should be valued as option value, not backlog, until task-order awards establish funded revenue and cash conversion. Its likely competitive set for scaled defense AI work includes PLTR, LDOS, BAH, and SAIC, which have incumbent relationships, cleared workforces, and balance-sheet capacity; this raises the risk that the vehicle broadens bid access without materially improving win rates. In the next 1-3 months, funded-order disclosures and operating-cash-flow trends matter more than contract-ceiling headlines; over 6-18 months, sustained losses would make dilution a more consequential catalyst than revenue growth.

The contrarian distinction is that INOD and BBAI should not trade as one "unloved AI" basket. INOD has a potentially measurable transition from labor-led revenue to product economics, whereas BBAI remains a government procurement-duration and financing-risk story. The asymmetric opportunity is therefore in INOD only after verification, while BBAI’s depressed equity price alone is insufficient evidence that future contract capacity is mispriced.

Key Decisions for Investors

• Place INOD on a conditional long watch for the next earnings release: initiate only if management discloses platform ACV/paid-customer conversion and consolidated gross-margin expansion while the largest-customer share continues declining. Target a 3-6 month rerating on recurring-revenue evidence; exit if the lead customer re-accelerates as a share of sales or growth decelerates without platform contribution.
• Express relative quality via long INOD / short BBAI in equal volatility-weighted notional over 3-6 months, rather than owning both outright. The trade isolates productization and improving concentration against uncertain IDIQ conversion and financing risk; cover the BBAI short if it reports material funded task orders plus a credible path to positive operating cash flow.
• Do not underwrite BBAI’s contract ceiling into forward revenue estimates until the company provides funded task-order values, expected timing, and associated gross-margin assumptions. Set an event alert for quarterly bookings-to-revenue conversion and cash burn; a capital raise before demonstrated conversion would be a negative catalyst.
• Monitor PLTR, LDOS, BAH, and SAIC defense-AI awards as a competitive read-through. Repeated large awards to incumbents without BBAI task orders over the next two quarters would falsify the thesis that procurement access alone creates meaningful share gain.