In short
Part-time US contractor role (20+ hrs/week, ~1 month with possible extension) evaluating AI-generated code for LLM training: building Python agents and tools to detect error patterns, judging model solutions for correctness, scalability, and production readiness, and benchmarking AI coding solutions. Core stack: Python plus full-stack experience (JS/React, C/C++, Java, Rust, or Go relevant).
About the role
Create and refine code examples and software solutions for large language model training and benchmarking. Review AI-generated code and judge whether it is correct, efficient, scalable, reliable, and ready for production.
- Build Python agents and automated tools that find error patterns and check software solutions.
- Assess model-generated work across prototyping, architecture, API design, production implementation, launch, experimentation, monitoring, and ongoing operations.
- Help develop AI coding solutions and evaluate them against industry performance benchmarks.
- Write clear, structured explanations for your evaluations.
- Location: Open to candidates based in the United States.
- Time: At least 20 hours per week.
- Engagement: Part-time contractor role planned for one month, with possible extensions based on performance and fit.
- Bring strong Python skills, including knowledge of production frameworks, development tools, and best practices.
- Have experience building full-stack applications and deploying scalable software, with knowledge of software architecture, development, debugging, and code review.
- JavaScript and ReactJS, C/C++, Java, Rust, or Go experience is relevant.
- Be fluent in English.
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training workAI training is the human work behind AI systems: people review examples and model answers so systems can learn from useful, accurate feedback. Software engineers bring the skills to assess generated code and build tools that check whether solutions work reliably.
Requirements
- Based in the United States
- Available at least 20 hours per week
- Strong Python skills, including production frameworks, development tools, and best practices