Research Engineer, Post-Training

Highlights

• Scale post-training loop turning expert feedback into better legal models.
• Push agent performance while balancing cost, latency and governance.
• Design grading systems strict enough for high-stakes legal work.

What they offer

• Equity
• Bonus

Ideal candidate

Self-directed researcher-engineer who geeks out on model behavior and agent traces, thrives in ambiguous applied research, and wants to turn expert feedback into models that matter for real legal work.

Requirements

• Hands-on experience with post-training or model-training work, such as SFT, preference optimization, RLHF/RLAIF, reward modeling, distillation, or adapting open-weight models to specialized domains.
• Strong judgment about model behavior: you can read traces, inspect outputs, identify failure modes, and reason about whether a metric is measuring the thing that matters.
• Strong Python and research-engineering ability. You can write clean code, debug experiments, and build the simple but reliable systems needed to make research move faster.
• Ability to self-manage ambiguous applied research projects and communicate clearly with researchers, engineers, product teams, domain experts, and external partners.