Research Engineer: Post-Training & Evaluation for Legal AI

Harvey; San Francisco (CA); On-site; USD 231,000 - 340,000; Full time

Job summary

Harvey is seeking a Research Engineer to drive model training experiments and improve agent performance in legal tasks. This role involves close collaboration with internal and external research partners to optimize training processes and develop efficient evaluation systems.

Ideal candidates will have hands-on experience with model training, proficiency in Python, and the ability to self-manage applied research projects. The compensation ranges from $231,000 to $340,000 annually.

Qualifications

• Hands-on experience with post-training or model-training work.
• Strong judgment about model behavior, capable of inspection and reasoning.
• Experience managing ambiguous applied research projects.

Responsibilities

• Drive post-training experiments, pushing agent performance.
• Optimize agent harnesses for improvements in quality.
• Design and develop reliable grading and reward systems.
• Study agent behavior and convert findings into training data.
• Collaborate with researchers to define and evaluate experiments.

Skills

Model training work

Python

Research engineering

Self-management

Tools

ML workflows

Experiment tracking

Job description

Harvey is seeking a Research Engineer to drive model training experiments and improve agent performance in legal tasks. This role involves close collaboration with internal and external research partners to optimize training processes and develop efficient evaluation systems.

Ideal candidates will have hands-on experience with model training, proficiency in Python, and the ability to self-manage applied research projects. The compensation ranges from $231,000 to $340,000 annually.