Research Engineer / Scientist, Reinforcement Learning

As a Member of Technical Staff, Modeling at Harvey, you will own the adaptation and improvement of leading open-weight AI models for complex legal reasoning and agentic workflows. You will design and execute post-training experiments using SFT and RL, focusing on reward design and verification for challenging legal tasks. This role offers a unique opportunity to shape research direction and see your work directly impact customers in a fast-paced, high-growth environment.

• Own research agenda for AI model capability in legal reasoning
• Design and run SFT/RL post-training experiments
• Develop reward functions and verifiers for complex tasks
• Debug and scale distributed training runs
• Build and utilize rigorous evaluation frameworks
• Hands-on experience post-training LLMs with SFT/RL