Vision-Language-Action (VLA) Annotator

We are seeking meticulous Vision-Language-Action (VLA) Annotators to contribute to the development of high-quality training datasets for vision-language-action and robotics models. This remote, full-time or contract position involves analyzing videos of people performing various real-world activities, segmenting these videos into meaningful action units, and creating or validating precise descriptions of the observed actions. Annotators will also review AI-generated annotations, correct errors, and ensure that the final data adheres to specific quality standards. Projects may cover a wide range of scenarios, including household, workplace, industrial, manipulation, and navigation activities.

Key Responsibilities
- Review videos thoroughly to understand the tasks or activities being performed.
- Segment videos into logical, action-based intervals with accurate start and end times.
- Capture all meaningful actions without gaps or overlaps in segmentation.
- Evaluate AI/VLA-generated captions, editing or replacing them as necessary.
- Write clear, concise descriptions based solely on visible actions in the video.
- Accurately describe relevant hands, objects, tools, movements, and locations when required.
- Maintain consistent terminology for objects and actions throughout each annotation.
- Identify and label invalid, idle, distracted, failed, recovery, or unusual video segments according to project guidelines.
- Review completed annotations for segmentation accuracy, caption clarity, grammar, consistency, and completeness before submission.
- Follow project-specific annotation guidelines and adapt quickly to changes in requirements.
- Flag ambiguous examples, recurring model errors, edge cases, or unclear instructions to the Operations or Quality teams.
- Utilize annotation shortcuts and AI-assisted tools efficiently while meeting quality and productivity targets.
- Participate in calibration, training, and feedback sessions as needed.

Required Qualifications
- Strong attention to detail with the ability to apply detailed written guidelines consistently.
- Proficient written English skills, capable of describing actions clearly using simple and precise language.
- Excellent observational skills to distinguish subtle differences in actions, objects, and timing.
- Comfortable reviewing video content for extended periods.
- Ability to perform repetitive annotation tasks while maintaining accuracy and consistency.
- Basic computer proficiency and quick adaptability to new annotation platforms.
- Ability to work independently and meet defined quality and productivity goals.
- Reliable computer setup and stable internet connection.
- Openness to feedback and ability to incorporate guideline changes promptly.

Preferred Qualifications and Benefits
- Previous experience in data annotation, video annotation, quality assurance, robotics data, computer vision, or AI training data.
- Experience with temporal segmentation or action recognition tasks.
- Familiarity with AI-assisted labeling tools or reviewing model-generated outputs.
- Experience working with egocentric or first-person video data.
- Basic understanding of robotics, manufacturing, household tasks, or other physical-world activities.

Success in this role is demonstrated by accurate segmentation of real-world actions, clear and consistent captions, minimal missed actions or timing errors, strong judgment in reviewing AI-generated annotations, and maintaining consistent quality across large volumes of video data. Additionally, the ability to identify recurring annotation or model issues rather than treating each task in isolation is highly valued.