Layer Health is a company founded by leading machine learning researchers from MIT and Harvard Medical School, focused on creating an AI layer to synthesize information from medical records. The role involves developing and iterating machine learning models for clinical use cases, conducting evaluations, and improving model performance using modern LLM techniques.
Responsibilities:
- Own end-to-end development of models for specific use cases: Build and iterate using our ML/LLM-powered workflows for new clinical areas, from initial prototyping through production refinement
- Conduct evaluation & error analysis: Own rigorous evaluation of model performance; perform deep error analysis to identify systematic failure modes and drive targeted improvements
- Ship frequent, high-quality updates to models based on data, feedback, and observed edge cases
- Leverage LLM application & tooling: Apply modern LLM techniques (prompting strategies, structured outputs, tool use, eval frameworks) to improve accuracy and robustness
- Support analytics & metrics: Contribute to metrics, reporting, and internal dashboards that track model performance and downstream impact