The work
You will evaluate coding agents that perform real software engineering work. Using familiar codebases and realistic tickets, you will judge whether an AI model's implementation works, identify errors, and explain how the output should be improved.
- Give an AI coding agent a real engineering ticket and review its implementation like a teammate's pull request.
- Run generated code and assess whether it solves the assigned problem correctly.
- Work step by step through software tasks in codebases you know.
- Pair with the AI on your own engineering work and identify failures in its reasoning or implementation.
- Provide clear technical feedback to improve future coding-agent performance.
What it pays and takes
This is a remote contract role for candidates in Mexico. The engagement may continue week to week when quality standards are met, and additional hours may be available.
- Pay: Up to $60 per hour, with weekly payments.
- Schedule: 40 hours each week, with the ability to set your own schedule.
- Location: Open to candidates in Mexico.
- Language: Strong written English.
- Experience: At least four years of professional software engineering experience.
- Education: Bachelor's or master's degree in computer science or a related field.
- Skills: Strong software engineering judgment, confidence running code, and the ability to explain why an implementation succeeds or fails.
- AI training experience is not required.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.