BioStack Platforms
Real world training envs for healthcare AI models
About BioStack Platforms
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.
Public traction evidence
Each signal links to the public source used for attribution.
- X
0-1M in <3 months Seven fig contracts locked in 21X growth 22 active clients LFGGGG! Quote Y Combinator @ycombinator · May 25 BioStack (@getbiostack) builds simulation environments where healthcare AI models practice on real clinical data. Messy records, lab tests, notes, outcomes unfolding over months— BioStack turns that into a post-training loop: data, evals, rewards, benchmarks. They've grown from 0:11 / 1:13
0-1M in <3 months Seven fig contracts locked in 21X growth 22 active clients LFGGGG! Quote Y Combinator @ycombinator · May 25 BioStack (@getbiostack) builds simulation environments where healthcare AI models practice on real clinical data. Messy records, lab tests, notes,...
- X
An AI model can get an ECG diagnosis right without reading the ECG. We tested frontier models from @OpenAI , @AnthropicAI , and @GoogleDeepMind , alongside smaller open models. After fine-tuning, models improved at predicting heart rate and electrical axis. But those values were Show more getbiostack.com Learning Around the Waveform: AI Loves Labels — BioStack
An AI model can get an ECG diagnosis right without reading the ECG. We tested frontier models from @OpenAI , @AnthropicAI , and @GoogleDeepMind , alongside smaller open models. After fine-tuning, models improved at predicting heart rate and electrical axis. But those values...
- X
Everyone is racing to build medical AI agents. Almost nobody is stress-testing the judges grading them. We tested frontier models as judges on @OpenAI HealthBench answers. Then we hit the answers with adversarial perturbations: Delete a negation. Tamper with clinical values. Show more getbiostack.com The Grader Is the Attack Surface in Medical AI — BioStack
Everyone is racing to build medical AI agents. Almost nobody is stress-testing the judges grading them. We tested frontier models as judges on @OpenAI HealthBench answers. Then we hit the answers with adversarial perturbations: Delete a negation. Tamper with clinical values....
- LinkedIn
Excited to share that we are joining Y Combinator!
Excited to share that we are joining Y Combinator! The founder says BioStack Platforms builds healthcare simulations for training AI models on clinical problems. Visible metrics: 407 likes, 407 reactions, 62 comments.
- X
@demishassabis and @garrytan in the house! Exactly 10 years after the AlphaGo moment, DeepMind is leading the charge on long/short horizon RL on all fronts. Post-training is going to eat the world. @ycombinator @GoogleDeepMind @GeminiApp Y Combinator and Garry Tan
@demishassabis and @garrytan in the house! Exactly 10 years after the AlphaGo moment, DeepMind is leading the charge on long/short horizon RL on all fronts. Post-training is going to eat the world. @ycombinator @GoogleDeepMind @GeminiApp Y Combinator and Garry Tan
- X
The SF hill climbing is the only hill climbing that matters. Please reinforce that learning.
The SF hill climbing is the only hill climbing that matters. Please reinforce that learning.
- X
@RuxandraTeslo @JackScannell13 I would still be bullish on AI to comprehensively map out the ‘avoidome’.
@RuxandraTeslo @JackScannell13 I would still be bullish on AI to comprehensively map out the ‘avoidome’.
- X
New merch just dropped @getbiostack
New merch just dropped @getbiostack