
What you’ll do: review paired AI-generated developer-facing writing (CLI help text, error messages, API reference documentation) against a client’s task brief and writing guide; read diffs and terminal output closely enough to catch a wrong technical explanation; select a preferred sample (or flag a tie / not suitable to judge); write a one-line reason and apply tags that support your verdict; clear a calibration gate against a known-answer test set before moving to live data.
Background: led developer documentation at a major tech company (cloud platforms, API-first products, developer tools); comfortable writing or evaluating CLI help, error messages, and API references; prior experience rating AI-generated writing is a strong plus.
Schedule: 5 hours/day, 5 business days for the pilot, ramping into an ongoing month-to-month engagement.
We consider all qualified applicants without regard to legally protected characteristics and provide reasonable accommodations upon request.
Mercor partners with leading AI labs and enterprises to train frontier models using human expertise. You will work on projects that focus on training and enhancing AI systems. You will be paid competitively, collaborate with leading researchers, and help shape the next generation of AI systems in your area of expertise.