AI Agent Task QC Engineer

You will review long-horizon tasks designed for AI agents before they are released for training and evaluation. Your focus is to catch incorrect assumptions, unclear instructions, missing edge cases, and tasks that do not reflect realistic workflows.

You will also test the grading rubrics and run tasks end to end in their connector environments. When something fails, you will reproduce and debug it, write precise findings, and share actionable feedback with task authors and connector engineers.

• Review agent tasks for realism, correctness, scope, and completeness.
• Pressure-test rubrics to ensure they define completion precisely and produce consistent grades.
• Check that tasks are achievable and that expected outcomes hold in connector environments.
• Reproduce failures and identify the cause.
• Help improve QC standards, checklists, and processes as the initiative grows.

What It Pays and Takes

The listing does not provide a compensation rate. This is a part-time contractor role listed at entry level, with a minimum weekly commitment and a required time-zone overlap.

• Pay: Compensation is handled through the project budget fields shown on the job page.
• Schedule: At least 4 hours per day and a minimum of 20 hours per week.
• Availability: At least 4 hours of overlap with PST each day.
• Location: Open to candidates in India, Pakistan, Nigeria, Egypt, Ghana, Bangladesh, Turkey, and Mexico.
• Language: Fluency in English.
• Technical skills: Strong backend software engineering skills, primarily in Python, including reading, running, and debugging code independently.
• Platform knowledge: Working knowledge of GCP, Docker, virtual machines, and Harbor.
• Work style: Exceptional attention to detail, strong judgment about real-world workflows, and precise written communication.
• Required tools: High daily proficiency with AI coding tools such as Claude Code, Cursor, or Copilot.

How It Works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI Training Work

AI training work is the human side of building artificial intelligence: people create, test, label, and evaluate examples that help AI systems learn. OpenTrain is a platform where people find these projects, build a profile, and apply in minutes.