Video Annotation Expert

Help AI understand what is happening inside a video

OpenTrain AI is recruiting Video Annotation Experts for a remote project focused on improving the visual data used to train and evaluate AI systems.

The work goes beyond simply labeling what's visible on screen. You'll review existing annotations, determine whether actions and hand positions have been captured correctly, identify inconsistencies and make precise corrections when the data doesn't meet project standards.

Those decisions contribute to datasets used for areas such as computer vision, action recognition and video understanding.

The project is expected to run for 3–6 months, with 30 contractor positions available and advertised compensation of $15–$20 per hour.

What you'll actually be reviewing

You'll work across different video datasets where consistency matters just as much as individual accuracy.

Assignments can include checking whether an action has been described correctly, validating hand-pose information and finding annotations that are incomplete or inconsistent with the project's guidelines.

Your responsibilities may include:

• Reviewing video annotations for accuracy, completeness and consistency.
• Detecting problems within action labels and hand-pose data.
• Correcting annotations that don't meet the required quality standard.
• Evaluating English-language action descriptions against established guidelines.
• Validating hand-pose information using detailed project criteria.
• Recording findings and corrections in standardized formats.
• Providing concise written explanations when annotation problems are identified.
• Suggesting improvements when existing guidelines create ambiguity.
• Delivering assigned work according to project deadlines and quality expectations.

Accuracy matters more than speed alone

This position is particularly suited to someone who can work methodically with large amounts of visual information without allowing small inconsistencies to slip through.

OpenTrain is looking for candidates with experience in video labeling workflows as well as knowledge of video understanding or computer-vision AI.

Familiarity with areas such as large-scale video datasets, action recognition, hand-pose analysis or similar annotation projects could be particularly relevant.

You'll also need to be comfortable working independently, interpreting detailed instructions and communicating issues clearly in written English.

What you'll need

Candidates should have:

• Experience with video labeling or annotation workflows.
• Knowledge of video understanding and computer-vision AI.
• Strong attention to detail.
• Fluent English.
• Clear written and verbal communication skills.
• The ability to consistently follow detailed annotation and quality standards.
• Good organization and independent working habits.
• The ability to work remotely and meet agreed deadlines.

Although this is an AI training project, previous employment specifically within the AI industry isn't required. What's more important is having the relevant video annotation expertise and being able to apply it consistently.

Who can apply?

Unlike some OpenTrain opportunities that are available globally, this project is restricted to candidates based in one of 28 eligible countries:

United States, United Kingdom, Mexico, Brazil, Spain, Türkiye, India, Sri Lanka, Germany, Netherlands, Finland, Iceland, Greece, Argentina, Philippines, South Korea, Japan, Australia, Austria, Belgium, Chile, Canada, Colombia, Czech Republic, Denmark, France, Italy, or Jordan.

Applicants must also be fluent in English.

A route into hands-on AI training work

For someone already familiar with video annotation, this project offers a relatively direct way to apply that expertise to modern AI development.

Instead of building the underlying computer-vision models, you'll work on an equally important part of the process: making sure the human-reviewed data those systems learn from is accurate and consistent.

The position is fully remote, runs for an expected 3–6 months, and pays an advertised $15–$20 per hour on a contractor basis.