Video Annotator
Mid-level Video Production role at The Global Talent Co.. Remote, Kenya · full-time.
Growth Roles summary, based on the employer's posting.
What you'll do
- Review each assigned clip fully, then divide it into clear action-based events
- Set precise beginning and ending timestamps for every video segment
- Write concise, grammatically correct instructions that match each visible action
- Apply project taxonomies and style rules while keeping annotations consistent
- Flag unusable footage, join calibration sessions, and meet quality and output targets
What you bring
- C1-level English or above, with strong grammar, vocabulary, and sentence construction
- Sharp visual attention and the ability to notice subtle scene changes
- Clear, concise writing that describes actions accurately
- Comfort using browser-based video and annotation tools after training
- Independence, consistency, basic computer skills, and reliable internet access
Who this fits
You’ll suit this role if you can work independently through repetitive visual tasks while following detailed rules closely. The position is fully remote and full-time, and you need a reliable internet connection plus a computer with a modern browser. You’ll also need to protect project data and respond constructively to calibration and quality feedback.
From the employer
Video Annotator
Location: Fully Remote Schedule: Full-Time, Remote
Check out The Global Talent Co.’s talent network: https://www.globaltalent.co/network
About the Role
We are looking for detail-oriented Video Annotators to support a Vision-Language-Action (VLA) dataset project. In this role, you will watch short video clips of physical tasks/activities and produce event-level annotations by segmenting each video into meaningful steps and writing clear, precise natural-language instructions describing the action taking place in each segment.
Key Responsibilities
Watch assigned videos in full before beginning annotation to understand the overall task and context.
Segment each video into discrete, logical steps/events based on visible action boundaries.
Write clear, concise, and grammatically correct step-by-step descriptions for each segment, accurately reflecting what is happening on screen.
Ensure annotations are action-oriented, unambiguous, and consistent in tense, tone, and phrasing across the dataset.
Assign accurate start and end timestamps for each segment.
Follow annotation guidelines, taxonomies, and style guides provided by the project team.
Flag videos that are unclear, corrupted, mislabeled, or otherwise unsuitable for annotation.
Participate in calibration sessions and incorporate reviewer/QA feedback to improve annotation quality and consistency.
Meet daily/weekly productivity and quality targets without compromising accuracy.
Maintain confidentiality of all project data and materials.
Required Qualifications
English Proficiency: C1 level (CEFR) or above — strong command of grammar, vocabulary, and sentence construction is essential.
Excellent attention to detail and ability to spot subtle changes in visual scenes.
Strong written communication skills; ability to describe actions concisely and precisely.
Comfort working with video-based tools and annotation software (training provided).
Ability to work independently, follow detailed guidelines, and maintain consistency over repetitive tasks.
Reliable internet connection and access to a computer with a modern browser.
Basic computer literacy (file handling, web-based tools, spreadsheets).
Preferred Qualifications
Prior experience in data annotation, labeling, transcription, or content moderation.
Familiarity with AI/ML concepts, especially computer vision, NLP, or robotics datasets.
Experience writing instructional or procedural content (how-to guides, SOPs, recipes).
Exposure to annotation platforms (e.g., CVAT, Label Studio, Scale, or similar tools).
At The Global Talent Co., our mission is to bridge the gap between leading global brands and amazing global talent. We believe everyone deserves equal access to opportunities, regardless of the country they are born in.
Not this one either?
Claude or ChatGPT reads the other 62,148 for you.
Get better matches →