Freelance Agent Evaluation Engineer
Freelance Agent Evaluation Engineer is an open role at Mindrift, available remotely, with pay listed as ≈ $40/hr.
Applications open on Mindrift’s own site. How we make money.
About this role
Please submit your CV in English and indicate your level of English proficiency. Mindrift connects specialists with project-based AI opportunities for leading tech companies, focused on testing, evaluating, and improving AI systems. Participation is project-based, not permanent employment. We're building a dataset to evaluate AI coding agents - how well a model handles real-world developer tasks. You'll create challenging tasks and evaluation criteria within realistic simulated environments: Build realistic developer environments - a virtual company with codebase, infrastructure, and context (tickets, docs, conversations) that forms a believable development history Design tasks from intermediate states of these environments - craft the prompt, define what "solved" means, and ensure the task is solvable by an AI agent Write tests that verify agent solutions - accept all valid approaches and reject incorrect ones, neither too strict nor too lenient Iterate on tasks and tests based on QA feedback - review agent solutions, analyze failures, and refine until the evaluation is fair and robust What this is NOT: Not data labeling Not prompt engineering Not writing code from scratch - the agent writes most of the code; you guide and evaluate What we look for: 5+ years in software development Core stack: Python (FastAPI), JavaScript/TypeScript (React), Docker, Postgres, Kafka, Redis Experience writing tests (functional, integration) English proficiency - B2+ Why this is hard: Frontier 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. Requirements and benefits Educational qualifications A Master’s Degree in Computer Science, Software Engineering, Data Science / Data
Pay
≈ $40/hr
As listed by Mindrift: “Your rate depends on the qualification tier you reach and how efficiently you complete tasks — up to the equivalent of $40/hr.”
About Mindrift
Mindrift's connection to Toloka: Toloka contributor work migrated to Mindrift. AI training, reasoning, coding, multilingual evaluation
Ownership confidence: Confirmed. See the full network profile →
Similar open roles
Frequently asked questions
How do I apply for the Freelance Agent Evaluation Engineer role at Mindrift?
Applications go through Mindrift's own portal - HumanSourcer links to the listing and never collects applications or handles hiring. Mindrift's access model is "Open / self-serve" (Open roles / applications). This listing was first seen here on September 3, 2026 and was still live at the last check.
Is the Freelance Agent Evaluation Engineer role at Mindrift remote?
Yes - Mindrift lists this role as remote.
What does the Freelance Agent Evaluation Engineer role at Mindrift pay?
Mindrift indicates ≈ $40/hr for this role, but hedges the figure rather than committing to it. Treat it as an indication: HumanSourcer flags hedged pay instead of presenting it as a firm rate, and the listing's own wording is quoted in the Pay section above.
What kind of work is Freelance Agent Evaluation Engineer?
AI model evaluation. AI training, reasoning, coding, multilingual evaluation
Is Mindrift a legitimate AI-training platform?
Toloka contributor work migrated to Mindrift - and that relationship is rated "Confirmed" here because it can be checked against mindrift.ai rather than taken on Mindrift's word. Confidence describes how well the ownership is evidenced. It is not a rating of how the platform treats the people who work for it, which no public source covers reliably.
Get new AI-training roles by email
One weekly digest of newly listed roles across every tracked network - pay where it's disclosed, no spam, unsubscribe anytime.
Free. See how we make money.