Beyond the labs: who buys expert data next

What the data says

Funding is the latest round I could confirm from the linked source. The "expert need" column is my hypothesis about who each company would need to check its models; treat it as a question for a discovery call, not a fact.

By vertical

Proof they already pay for experts

Size one segment

A bottom-up model for expert evaluation spend. Inputs marked "assumption" are mine and you should change them. The default panel size comes from HealthBench (262 physicians), a real program.

Who could bring these buyers in

Selling one account at a time is slow. These groups touch many of the buyers above at once, and each has its own reason to want a reliable source of experts.

Shared investors

Run your own query

The whole dataset is a SQLite database running in your browser (sql.js). Pick a preset or write your own. Tables: accounts, investors, evidence, vendors, market, multipliers.

How I'd test this in 30 days

1. Week 1Pick five. Take the five health accounts where the expert need is clearest. Find who owns clinical evaluation at each one (clinical AI lead, head of clinical, medical director).
2. Week 2Five discovery calls. Ask how they find clinician reviewers today, what a round costs, how long it takes, and what breaks. Replace my "expert need" guesses with what they say.
3. Week 3One paid pilot. Offer a fixed-scope evaluation sprint, priced per completed rubric, with a turnaround the buyer can't get from its own clinicians.
4. Week 4Decide. Compare pilot margin and sales effort with lab work. If the numbers hold, write the playbook and take it to the legal accounts next. If they don't, say so and stop.

Method and sources

• Accounts were picked as well-funded vertical AI companies whose products need expert judgement to check. It is a sample for testing a thesis, not a full market list.
• Funding figures are the latest round confirmed in the linked source as of 30 Oct 2026. Some outlets report later talks; those are left out unless a round was announced.
• Vendor figures mix gross run rates and annual revenue across different periods. They are compared with the market estimate only to show the order of magnitude.
• Nothing here comes from private data. No company named on this page reviewed or endorsed it.

All sources