Owning end-to-end coding data projects, the full-time Research Engineer, Coding Evaluation will design and run systems that teach frontier models how to code, focusing on training data quality and evaluation in a remote environment.
Key responsibilities:
Manage coding data projects from initial design through execution and scale-up
Design training workflows and evaluation rubrics that reflect real-world software engineering tasks
Collaborate with clients and internal teams to enhance coding data products and technical processesRequired qualifications 3-6+ years of professional software engineering experience
Strong coding ability in at least one mainstream programming language
High standards for engineering quality and correctness
Experience with debugging technical environments, including containers and automated test setups
Interest in AI/ML systems and their impact on coding capabilities
