AI Coding Agent Evaluation Specialist at toloka-ai

Mindrift is recruiting experienced software developers for a project-based freelance opportunity to create and evaluate realistic coding tasks and automated tests used to assess AI coding agents. The work involves simulating developer environments, defining solvable tasks and evaluation criteria, writing functional and integration tests, reviewing agent solutions, and refining tasks based on QA feedback. The role is remote and part-time, paid per accepted task at a rate of up to the equivalent of $50 per hour, and is not permanent employment.