AI Trainer (SWE)

• Create challenging technical tasks.
• Evaluate model-generated code.
• Identify failures.
• Establish objective standards for correct solutions.
• Create realistic software engineering tasks that challenge advanced AI models.
• Develop task instructions, reproducible environments, reference solutions, tests, and evaluation criteria.
• Evaluate model-generated code and technical deliverables for correctness, robustness, and completeness.
• Identify issues involving implementation, debugging, dependencies, performance, testing, or reproducibility.

• Prior paid experience training, evaluating, benchmarking, or improving AI systems.
• Strong practical proficiency with Python.
• Professional experience with at least one additional programming language, such as Java, JavaScript, TypeScript, Rust, C++, C#, or Go.
• Strong knowledge of software design, debugging, testing, version control, and dependency management.
• Provide evidence such as substantial GitHub or GitLab repositories, meaningful open-source contributions, maintainer or core-contributor responsibilities, Codeforces, ICPC, Olympiad, or other programming-competition results, public developer tools, libraries, systems, or technical projects, or technical publications, benchmarks, or engineering write-ups.

• Experience creating AI-training tasks, reference solutions, tests, graders, or benchmark problems.
• Experience reviewing other AI trainers’ technical submissions.
• Codeforces rating of approximately 1900 or higher, or strong results in another programming competition.
• Experience with Linux, Docker, CI/CD, command-line tools, or sandboxed environments.
• Prior experience with long-horizon or agentic coding evaluations.
• Availability to begin immediately.

Project and contracting process: the hiring partner. Applications continue on the provider's website.