BioStack, a San Francisco startup from Y Combinator, announced $9.5 million in seed funding on Thursday, Oct. 22, to turn more than 100 million de-identified patient records into training data for medical AI.
Why it matters: AI labs want models that can reason through real patient cases, not just pass medical exams. Hospital records that could teach them are scattered across systems, and BioStack is trying to stitch them together.
Who invested
Peak XV, Afore Capital, DCVC, Emerson Collective, SV Angel, Verdict Capital, Streamlined Ventures and Y Combinator backed the round, the company said.
What it does
- Connects notes, labs, imaging, medications, treatments and outcomes from hospitals in the U.S., India, Brazil and Argentina.
- Builds practice tasks for AI models, such as ordering the next test or adjusting a medication, then scores the model against real outcomes, according to its Y Combinator profile.
- Strips identifying details from the records before they go into datasets, the company says.
BioStack was founded in 2025 by CEO Sanat Mishra and CTO Parth Patwa, Digital Health News reported. Its YC profile lists a team of 10 in San Francisco.
The money will go to expanding in the U.S. and abroad and hiring engineers and researchers, the company said.
More SF AI funding: Town hits a $1B valuation.