An application made for this job — a tailored resume and cover letter that speak straight to the posting.
Braintrust is seeking a production software engineer to review AI-generated code, diagnose root causes and craft reference solutions for AI training. This is a software engineering and code-evaluation role, not manual testing.
You will review correctness, reliability, security, performance and maintainability across unfamiliar codebases, identify architectural weaknesses and propose improvements. Strong debugging, English communication and collaboration with researchers are essential.
Apply your production engineering judgment to improve AI-generated code. You will evaluate implementations, identify root causes and develop clear reference solutions for AI training. This is a software engineering and code-evaluation role, not a manual testing position.
Expected project length is three months. Availability of at least 40 hours per week, 8 hours per day and four hours of daily Pacific Time overlap is required. Selection includes an approximately 20-minute client AI interview and a 45-60-minute technical delivery interview. Include your LinkedIn profile URL in your required resume.