AI Agent Task QC Engineer

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. OpenTrain AI hires and contracts contributors for projects that help develop and evaluate modern AI systems, giving freelancers a place to build a lasting portfolio of specialized work.

• Remote contractor opportunity
• Part-time schedule of at least 20 hours per week
• Open to candidates in India, Pakistan, Nigeria, Egypt, Ghana, Bangladesh, Türkiye, and Mexico

About AI Agent Training

AI training is the human side of building artificial intelligence. Contributors create, review, and evaluate examples that help AI systems understand instructions, complete workflows, and produce reliable results. In this role, your quality reviews help ensure agent tasks and evaluation criteria reflect realistic use cases.

• Work on cutting-edge AI agent training and evaluation
• Use technical judgment to identify errors, ambiguity, and missing edge cases
• Help improve how AI systems are tested against real-world workflows

The AI Agent Task QC Engineer Role

OpenTrain is seeking an AI Agent Task QC Engineer to support AI Agent Task QC Review. You will join a quality-control team working on a frontier AI data initiative for training and evaluating AI agents. As a final quality check before tasks ship, you will determine whether tasks and rubrics are realistic, correct, achievable, and unambiguous.

• Role type: Part-time contractor
• Experience level: Entry level
• Primary language: English
• Subject area: AI Agent Task QC Review

What You'll Do

You will review long-horizon agent tasks authored by a mining team and validate them from end to end. The work combines backend engineering, software debugging, rubric evaluation, and careful reasoning about whether tasks faithfully represent real-world workflows.

• Review agent tasks for realism, correctness, appropriate scope, flawed assumptions, ambiguity, and missing edge cases.
• Pressure-test rubrics to confirm they define completion precisely, support consistent grading, and cannot be gamed or misread.
• Validate tasks against the connector environments where they run.
• Verify that stated objectives are achievable and that expected outcomes hold.
• Reproduce and debug failures, then provide clear, specific findings to task authors and connector engineers.
• Define and improve QC standards, checklists, and processes as the initiative scales.

Required Skills and Qualifications

This role is listed at the entry level, but it requires strong practical technical ability and a high standard for correctness. You should be comfortable investigating code and environments independently, communicating precise findings, and judging whether an automated workflow is realistic.

• Strong backend software engineering skills, primarily in Python.
• Ability to read other people's code, run it, and debug it independently.
• Working knowledge of GCP, Docker, virtual machines, and Harbor.
• High daily proficiency with AI coding tools such as Claude Code, Cursor, or Copilot. This is a hard requirement.
• Exceptional attention to detail and the ability to find failures others missed.
• Sharp written communication for precise, actionable feedback on tasks and rubrics.
• Ability to reason about real-world workflows and evaluate whether a task faithfully represents one.
• Fluency in English.

Schedule, Availability, and Compensation

The expected commitment is at least 4 hours per day and a minimum of 20 hours per week. You must be available for 4 hours of overlap with Pacific Standard Time (PST). Compensation is handled through the OpenTrain project budget fields displayed on this job page.

• Minimum availability: 20 hours per week
• Minimum daily availability: 4 hours
• Required overlap: 4 hours with PST
• Compensation: See the project budget fields on this job page

Why Build an AI Training Career with OpenTrain

AI training and data-labeling work is a fast-growing way to work in tech. Modern AI systems depend on people who can prepare examples, evaluate outputs, and improve the quality of model behavior. This role offers the chance to contribute directly to how advanced AI agents are tested and developed.

OpenTrain helps freelancers discover AI training opportunities, build a profile, and develop credible proof of work. Creating an OpenTrain account is free, and your growing profile can support a longer-term career in AI training and data labeling.

• Remote work that can fit around other commitments
• Hands-on exposure to advanced AI agent evaluation
• A portfolio-building opportunity in a rapidly growing field
• Free OpenTrain account creation