Research Engineer, Reward Models Platform

The responsibilities are:

• Design and build infrastructure for researchers to rapidly iterate on reward signals.
• Develop automated systems to assess the quality of rewards and detect reward hacks.
• Build pipelines and workflows to reduce toil in reward development.

Required skills include:

• Strong Python skills and experience with ML workflows.
• Comfort working across the stack, from data pipelines to user-facing tooling.
• Ability to balance building robust systems with moving quickly in research.

The job also includes:

• Optimizing systems for performance, reliability, and ease of use.
• Contributing to best practices and documentation for reward development.
• Collaborating with researchers to translate science requirements into platform capabilities.