ppMindrift connects specialists with project-based AI opportunities for leading tech companies, focused on testing, evaluating, and improving AI systems. bParticipation is project-based, not permanent employment. /b /p pbWhat this opportunity involves /b /p pWe're building a dataset to evaluate AI coding agents - how well a model handles real-world developer tasks. /p pYou’ll create challenging tasks and evaluation criteria within realistic simulated environments: /p ulliBuild realistic developer environments - a virtual company with codebase, infrastructure, and context (tickets, docs, conversations) that forms a believable development history /liliDesign tasks from intermediate states of these environments - craft the prompt, define what "solved" means, and ensure the task is solvable by an AI agent /liliWrite tests that verify agent solutions - accept all valid approaches and reject incorrect ones, neither too strict nor too lenient /liliIterate on tasks and tests based on QA feedback - review agent solutions, analyze failures,



and refine until the evaluation is fair and robust /li /ul pbWhat this is NOT /b /p ulliNot data labeling /liliNot prompt engineering /liliNot writing code from scratch - the agent writes most of the code; you guide and evaluate /li /ul pbWhat we look for /b /p ulli5+ years in software development /liliCore stack: Python (FastAPI), JavaScript/TypeScript (React), Docker, Postgres, Kafka, Redis /liliExperience writing tests (functional, integration) /liliEnglish proficiency - B2+ /li /ul pbWhy this is hard /b /p pFrontier models are already good at coding. Creating a task that genuinely challenges the best models is non-trivial. You need to deeply understand where models fail and what scenarios reveal the difference between a good and a bad solution. Tasks have many valid solutions - writing tests that accept all correct solutions and reject incorrect ones is harder than it sounds. /p pbEffort estimate /b /p pTasks for this project are estimated to take 20 hours to complete, depending on complexity. This is an estimate and not a schedule requirement; you choose when and how to work. Tasks must be submitted by the deadline and meet the listed acceptance criteria to be accepted. /p pbCompensation /b /p pUp to b$40/hr equivalent /b, depending on level and pace. Tasks are estimated at ~20 hours each; you set your own schedule. /p /p #J-18808-Ljbffr

📌 Freelance Agent Evaluation Engineer (Barcelona)
🏢 Mindrift
📍 Barcelona