Research Engineer / Scientist, Reinforcement Learning

The Research Engineer / Scientist, Reinforcement Learning role at Harvey focuses on adapting and improving open-weight models for legal work through post-training techniques like SFT and RL. The position involves designing reward functions, running large-scale training experiments, and developing rigorous evaluations to enhance model capabilities in legal reasoning and agentic workflows. This role requires strong engineering and ML fundamentals to drive research from experiments to customer-facing models.