About the role
Braintrust is seeking a senior software engineer to apply production engineering judgment to AI-generated code. You will review implementations, diagnose bugs, and develop clear reference solutions for AI training, working across languages and architectures.
The role emphasizes maintainable design, security, and performance in production software. You will collaborate with researchers, craft evaluation rubrics, and communicate findings clearly in English.
Review AI-generated code for correctness, reliability, security, performance, scalability and maintainability. Diagnose bugs, incomplete implementations, edge cases and architectural weaknesses in unfamiliar codebases. Evaluate fixes, features, refactoring, APIs, configurations and database changes, comparing alternative solutions. Improve implementations and explain technical issues and recommendations clearly. Develop evaluation criteria, annotations and rubrics, collaborating with researchers on coding benchmarks. 7+ years of professional software engineering experience building, maintaining, debugging and reviewing production software. Strong proficiency in at least one programming language such as Python, JavaScript/TypeScript, Java, C++, Go, C#, Ruby, PHP or Rust. Strong knowledge of clean code, modular architecture, abstraction, error handling and software design. Experience identifying functional, performance, security and design issues, with strong debugging and root-cause analysis. Familiarity with Git, collaborative code reviews and modern engineering practices. Strong knowledge of data structures, algorithms, APIs, databases and application architecture. Strong written English and clear technical communication.