About CohortLoom: synthetic patients for health AI
CohortLoom makes test data for software that reads clinical records, where every answer is known in advance.
Why we built it
Teams that build clinical NLP, coding tools and EHR integrations keep running into the same wall. Real records are slow to get, carry heavy privacy obligations, and still need to be labelled by hand before they can tell you whether your software is right.
We think test data should come with its answers. So we start from a simulated patient's chart, plan each document before it is written, and ship the plan as an answer key alongside the text.
How we work
- Synthetic by construction. No real patient data, and no de-identified data, is used anywhere in making a dataset.
- Clean model lineage. Our writing models are open-weight models we run ourselves. They are never trained on Claude or Gemini outputs.
- Honest claims. We publish quality figures only with the validation report that backs them. Until then, the report and a sample are available on request.
- AI-operated, human-backed. Conway, an AI agent, runs our sales and support. A person at Apply AI Inc. is responsible for the business and reachable at [email protected]. Read the AI disclosure.
Company
CohortLoom is a product of Apply AI Inc.
Apply AI Inc.3845 McCoy Dr, Ste 101B
Aurora, IL 60504
United States
[email protected]