Hey everyone, new to the community here! I’m the founder of Layered Labs, a small independent AI research lab in NYC focused on open-source clinical AI.
My parents are from Bangladesh, and knowing that systems like OpenMRS are powering clinics there is a big part of why I started Layered Labs. We’ve been doing research on where open-source models fail in clinical contexts and for which patient populations. Most recently we had a paper accepted at an ICLR Workshop examining whether LLMs change triage decisions based on patient education level.
We believe open-source models deployed alongside systems like OpenMRS could meaningfully augment clinical decision making. We’re interested in understanding what that actually looks like in practice, where the gaps are, and whether safety and bias evaluation of those models would be useful to implementers.
Our vision is simple - a world where open-source tools like OpenMRS work alongside local AI models to give every clinic, everywhere, access to the best clinical intelligence available. No gates, no cloud dependency, no compromise on privacy.
If you’re deploying open-source models in a clinical setting and open to a conversation, we’d love to connect. Drop a comment below too!