Code-Data Eval Author — Software Eng at Mercor

Code-Data Eval Author (Software Engineer) | $35–120/hr | Remote — Americas & Europe

Mercor partners with frontier AI labs to build the evaluations that coding models are trained and measured against. As an Eval Author, you'll craft real engineering tasks, golden solutions, and test verifiers — defining what correct means for the next generation of AI coding models.

What You'll Do

• Author non-trivial coding tasks complete with golden solutions and automated verifiers
• Design grading rubrics and evaluate AI agent trajectories and model outputs
• Improve task and rubric quality through structured peer review

Who You Are

• 5+ years of software engineering experience at a product-focused organization (big tech or venture-backed startup)
• Strong fundamentals in code quality, systems design, debugging, and testing
• Clear, precise written communication — you write instructions others can follow
• Familiarity with AI coding tools or evaluations is a plus, not a requirement

Engagement & Compensation

• Remote contract with flexible 30+ hours per week
• Hourly rate scaled to your local market: US/Canada $90–120/hr; Europe and LatAm rates adjusted to region

Hiring Process — Paid

The process includes a Mercor Technical Screen, a live Code Review Session, and a Domain Expert Interview. You are paid $200 upon completing all three steps, regardless of the outcome.