Research
Research Engineer, Post-Training & Agents
New York, NYFull-time
Build the post-training and agentic systems that turn raw models into reliable medical operators. This role sits at the boundary between frontier research and product deployment.
Responsibilities
- •Develop fine-tuning, preference optimization, and tool-use pipelines for medical assistant behavior
- •Build agent scaffolds that can navigate records, care protocols, scheduling, logistics, and follow-up execution
- •Partner with product and clinical teams to define realistic tasks, benchmarks, and failure modes
- •Instrument experiments to measure quality, latency, cost, and safety tradeoffs
- •Ship research infrastructure that accelerates iteration across the applied research team
Requirements
- •Strong software engineering ability with experience in modern ML systems
- •Track record working on post-training, agent systems, evaluation, or model behavior improvement
- •Excellent Python skills and comfort working close to production systems
- •Ability to rapidly prototype while maintaining scientific rigor
- •Interest in building safe, high-reliability AI for real-world healthcare environments