YC-Backed BioStack Raises $9.5 Mn To Turn Hospital Records Into Medical AI Training Data

Sudip Kumar

Published on:

Follow Us

BioStack, a San Francisco-based healthcare data startup, has raised $9.5 million in a seed round led by Peak XV, DCVC and Afore Capital. Afore Capital, DCVC, Emerson Collective, SV Angel, Verdict Capital, Streamlined Ventures and Y Combinator also participated. The company did not disclose its valuation.

The startup turns scattered hospital records into datasets used to train and evaluate medical AI models. The money will go into expansion across the US and international markets, development of its technology platform, and hiring for engineering and research.

Read More: Pronto Founder Anjali Sardana Sets Up Physical AI Venture Provision Labs In The US

Read More: Unacademy Cofounders Gaurav Munjal, Roman Saini Raise Rs 700 Cr AI And Deeptech Fund

What BioStack Actually Builds

Patient information in most health systems sits in separate places. Doctors’ notes live in one system, lab results in another, scans in a third. None of them talk to each other in a way a model can use.

BioStack pulls these sources into longitudinal records that follow a patient over time. The record links clinical notes, laboratory results, medical imaging, medications, treatment histories and outcomes into one timeline.

Once de-identified, that data becomes the training material for medical research, drug development and healthcare AI. The company lists uses such as disease modelling, biomarker discovery, patient stratification and analysis of how patients respond to treatment.

The Data Network Behind the Pitch

BioStack says its network spans the US, India, Brazil and Argentina, with access to more than 100 million de-identified patient records. It works with hospitals in all four countries.

The composition matters. The company claims to have combined several hundred thousand X-rays, MRI and CT scans with several million telehealth audio transcripts, plus pathology and other clinical records. Multimodal data of that shape is what clinical AI agents need, because real diagnosis rarely comes from a single test.

Each provider system has its own data privacy protocols, and BioStack says its platform includes workflows built around those rules rather than a single global pipeline.

Why Hospitals Sell Data at All

Health systems sit on records they cannot easily use. Storing them is a cost, research access is slow, and most hospitals have no in-house team to structure the data for AI.

BioStack’s offer is to do that work and share the value. For a provider, it converts dead storage into a research partnership. For a model developer, it removes the hardest part of medical AI, which is not compute but clean, longitudinal, legally sourced patient data.

That is also where the risk sits. De-identification is a technical claim that regulators now test, and a re-identification incident in any of the four markets would end the business model quickly.

The India Link: Cancer Vaccine Programme

In India, BioStack is helping connect cancer patients with an investigational cancer vaccine programme. Participating patients’ clinical information supports the development work of biotech companies, and eligible patients may get access to personalised investigational treatments.

The company did not name the biotech firm or the vaccine. That silence is normal at this stage, but it also means the arrangement cannot be verified beyond the company’s own statement.

Who Is Behind It

BioStack was founded in 2025 by Sanat Mishra, who is chief executive, and Parth Patwa, who is chief technology officer. The company was part of Y Combinator’s Spring 2026 batch and has a team of 11.

Mishra conducted cancer genomics research at Stanford, Yale, Carnegie Mellon and the Max Planck Institute. He also worked with AI labs as an early researcher on reinforcement learning tasks in healthcare and biotechnology. He holds degrees from IISER Mohali and Carnegie Mellon.

Patwa is a former generative AI scientist at AWS and a former researcher at MIT. He has published more than 40 papers with over 1,800 citations at venues including CVPR, NeurIPS and EMNLP, and holds a master’s degree from UCLA.

What to Watch

The healthtech market is funding fewer companies with more money. Healthtech startups globally raised $22.8 billion across 1,150 rounds in 2026 so far, according to Tracxn. The same period of 2025 saw $19.9 billion across 1,530 rounds. That is about 15% more capital in roughly 25% fewer deals.

BioStack is on the right side of that shift, but three things will decide whether the round was a start or a peak. The first is a named hospital group or biotech customer. The second is whether the datasets hold up in published model benchmarks, not just in a pitch deck. The third is consent. A business built on other people’s medical records lives or dies on how it answers that question in each country it enters.