Vertical AI Software

This has been our highest-conviction category — where we concentrate the most capital, and where the compounding-loop argument becomes most concrete.

Pre-AI, building vertical software for a $500M niche took the same 100+ engineers as building for a $10B horizontal market. So most verticals went unserved, or got generic horizontal tools awkwardly bolted onto industry-specific workflows. AI collapses that cost structure — a 3-person team can now build a deep vertical product that would have taken 30 engineers two years ago. That collapse has unlocked hundreds of verticals that were uneconomic in the SaaS era, and the founders moving fastest into them now will be the incumbents everyone else has to unseat.

But the cost collapse isn't what makes vertical AI defensible — and this is the distinction most investors miss. Depth is. We look for verticals with genuinely hard problems: deep regulatory complexity, multi-stakeholder workflows with conflicting incentives, edge cases that multiply rather than converge. A company serving dental practices has to verify insurance eligibility before scheduling, cross-reference CDT codes against each payer's coverage matrix, and handle the cascade when a patient's plan changes mid-treatment. A commercial roofing platform has to know that a change order triggers a lien waiver, that requirements differ by state and municipality, and that the insurance implications shift with the sub's coverage. That depth is the test.

There's another dimension generative AI unlocks, and it's especially powerful down-market. Many of the deepest verticals serve owners with no IT department who were never going to hunt and peck through a SaaS app, no matter how many AI features it had. AI-native vertical companies can now offer a full-stack solution where the complexity is masked entirely by agentic front-ends — the owner describes what they need and the system handles it. Down-market, these aren't headless infrastructure; they become the complete solution, eventually an agentic model that runs the business for the owner. That's not a tool. It's a new relationship between a business and the technology that operates it.

The loop goes nuclear in verticals for two reasons. First, the TAM staircase. The best vertical companies don't stay at the point-solution level — they climb: point solution ($50–200/mo per seat) → all-in-one plus payments ($500–2K/mo plus processing) → labor replacement ($2–5K/mo, outcome-based) → business-in-a-box (a percentage of revenue). Each step generates richer traces, builds deeper judgment, and makes the system harder to replace. At the top, the AI doesn't help run the business — it runs the business. The founder coaches the system; the system operates the company. No one switches away from that.

Second, cross-node intelligence across supply and demand chains. The deepest verticals serve an ecosystem — supplier and buyer, contractor and client, provider and payer. When a platform gets agents operating on both sides of a transaction, the intelligence it accumulates isn't about one company, it's about how the chain itself works. That cross-node graph is structurally exclusive — no single participant could build it, and leaving means losing access to the collective intelligence about how your counterparties operate. The verticals where both dynamics are present are the highest-conviction investments in our portfolio.