The work
Create and improve software engineering examples used to train and evaluate large language models. Review AI-generated code for correctness, architecture, efficiency, scalability, and reliability, especially in systems and infrastructure work.
- Write, correct, and refine code in Python, C/C++, Rust, Go, Java, and JavaScript, including ReactJS.
- Build agents to verify infrastructure and systems code, and design automated checks that compare solutions with industry performance benchmarks.
- Identify recurring code errors, assess model abilities across software engineering tasks, and explain technical evaluations clearly.
What it pays and takes
- Location: Candidates must be based in the United States.
- Schedule: Less than 20 hours per week; the engagement lasts one month, with possible extensions based on performance and fit.
- Experience requirement: At least three years of software engineering experience.
- Bring strong systems programming, infrastructure, or backend development skills, especially in Python, C/C++, Rust, or Go.
- You should have experience building and deploying scalable production software, and a strong understanding of software architecture, development, debugging, and code review.
- Fluent English and the ability to give clear, structured reasons for technical judgments are required.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.