Description
Mindrift is seeking experienced software developers for a project-based freelance opportunity creating and evaluating benchmark tasks for AI coding agents. The role involves building realistic simulated developer environments, designing challenging coding tasks and success criteria, writing functional and integration tests, and refining evaluations based on QA feedback. The core stack includes Python with FastAPI, JavaScript/TypeScript with React, Docker, PostgreSQL, Kafka, and Redis. Participation is remote and part-time, with payment per accepted task up to the equivalent of $40 per hour; it is not permanent employment.
