Projects
Internist-7B: A 7B-parameter medical large language model and the first model of its size to surpass 60% on MedQA. Trained on a curated mix of high-quality medical and general-domain data, Internist-7B was designed to study how data composition and scale shape clinical reasoning in compact models. It serves as the research foundation for the clinical agent now deployed at the Cliniques Universitaires Saint-Luc.
Chopper (Internist.AI): An internal LLM-serving and orchestration stack powering the clinical agent deployed inside the Epic electronic health record at the Cliniques Universitaires Saint-Luc. Chopper handles model serving, inference routing, safety guardrails, auditability, and monitoring for an agentic system used by 3000+ clinicians across 150,000+ clinical interactions.