Fuse raises $25M to disrupt aging loan origination systems used by U.S. credit unions

Fuse raised a $25M Series A led by Footwork, Primary Venture Partners, NextView Ventures and Commerce Ventures. The AI-native loan origination system already serves 100+ customers and has set aside a $5M 'rescue fund' and free access for the first 50 qualifying credit unions to ease migration from legacy LOS vendors. Fuse positions its AI agents to automate underwriting, increase loan volumes and cut operational costs, targeting incumbents like nCino and MeridianLink.

Analysis

An AI-native LOS reduces the economic friction of switching by compressing manual workflows and pushing the marginal cost of underwriting toward software-defined agents. That dynamic creates a multi-year erosion risk to incumbents’ renewal economics rather than an immediate revenue cliff — expect early signs in decelerating net-new ARR and rising discounting pressure over the next 2–4 quarters as pilots convert.

Second-order winners include cloud/LLM infrastructure and middleware vendors that can capture recurring inference and data-ops spend; conversely, integrated legacy vendors with high fixed-cost service models will see margin compression as new entrants monetize via variable, usage-linked pricing. Increased processing throughput also creates credit-cycle feedback: if AI materially shortens origination time and approval friction, credit supply can expand faster than risk controls evolve, elevating loss-rate tail risk on a 12–36 month horizon.

Key reversal risks are regulatory and model-risk events — supervisory guidance on automated underwriting or a high-profile loss event could freeze deployments and reset vendor valuations sharply. Monitor concrete metrics from pilots (cost-per-loan, turn-time, realized loss on AI-originated loans) and legacy-vendor client renewal cadence; a single large credit-institution conversion or a notable regulatory bulletin in the next 3–9 months will be the fastest catalyst to re-rate this sub-sector.

Key Decisions for Investors

• Short NCNO via a 6–12 month put spread (buy puts / sell lower strike to fund) sized to capture a 30–50% downside if renewal visibility weakens; cap worst-case loss to the net premium (~20% of notional). Hedge with an offsetting long position in cloud infra if you need convexity protection.
• Pair trade: short NCNO (size A) / long AMZN or MSFT (size B) over 9–18 months — if LLM-hosting demand accelerates, cloud vendors capture durable margins while legacy LOS faces churn; keep aggregate beta neutral and cut pair if NCNO posts <3% QoQ ARR attrition or if cloud guidance disappoints.
• Opportunistic long JKHY or FIS (12–24 month horizon) small allocation to capture payments/core-processor upside as banks and credit unions modernize front-office stacks; limit exposure to 2–4% portfolio and exit if vendor partnerships with AI-LOS players fail to materialize within 12 months.
• Event hedge: buy short-dated NCNO puts ahead of the next earnings and guidance call as an inexpensive insurance against early disclosure of contract churn or pilot failures; if no negative surprise, roll or sell into the event-related volatility decay.