platoseedplatoseed
Real world training envs for healthcare AI models
BioStack is building the data engine for healthcare and drug discovery AI. The bottleneck is not models. It is access to high-quality biological data. Clinical and experimental data is fragmented, unstructured, and locked inside hospitals, labs, and CROs, while generating new data is slow and expensive. BioStack fixes this with proprietary clinical and preclinical data pipelines that turn real biomedical workflows into ML-ready training environments. We structure longitudinal multimodal data across imaging, EHR, and experimental assays, then package it for post-training and reinforcement learning so models can learn how research and care actually happen. Instead of static datasets, BioStack gives AI labs workflow-aligned data and environments that improve reasoning, decision-making, and real-world performance in biology and medicine.
BioStack Platforms provides curated, ML-ready healthcare datasets and data infrastructure to enhance AI models in real-world clinical contexts. They offer domain-specific data sources and tools to annotate, structure, and use healthcare data for training and evaluation.
BioStack sources and structures high-quality ML-ready healthcare data across EHR, labs, imaging, ECGs, notes, treatments, and patient outcomes. They provide novel pre-clinical and medical datasets, causal inference capabilities, and data points for pre/post training. The platform supports building RL environments for post-training, multi-agent reasoning infrastructure, and tools to annotate novel or public data, enabling users to fine-tune and deploy industry-grade AI models with domain-specific data and tailored reward functions for RL tasks.
Who it’s for: Teams developing healthcare AI models (e.g., researchers, data science groups, ML engineers) who need real-world clinical data and infrastructure to train, validate, and deploy AI models.
marketing/consultation call to engage; signs of productized data assets and RL environment capabilities
I'm the co-founder and CEO of BioStack Platforms. In another life, I was a cancer genomics researcher at Stanford, Yale, Carnegie Mellon, and the Max-Planck Institute. I also worked as an early researcher with top AI labs on RL tasks and benchmarks across healthcare and biotech. Hailing from Mumbai, I attended IISER Mohali and Carnegie Mellon for my BS and MS degrees respectively.
AmazonI am the Co-founder and CTO at BioStack Platforms. Prior to this, I was a Gen AI scientist at AWS. Before that, I was a researcher at MIT. I have published 40+ papers with 1800+ citations across top venues like CVPR, neurIPS, EMNLP etc. in ML, GenAI, NLP. I completed my MS from UCLA.
We give AI labs the full post-training loop for healthcare
BioStack Platforms creates realistic healthcare simulation environments using real patient data so AI labs can train and test medical models before deployment. The environments include clinical tasks, feedback loops, and reward functions for model improvement.
▲ 8
Multimodal foundation models to predict uncollected patient biology
Multimodal data provider for robotics and world modeling
