32
Put your video captioning skill to work training frontier AI. Correct highly detailed, structured descriptions of short video clips so they match the footage exactly.
What You’ll Do
Work from a provided first-draft caption and correct it against the footage. Set shot boundaries to the exact frame, classify transitions and camera moves, transcribe dialogue with speaker attribution, and log actions, sounds, and on-screen text with timestamps.
Who We’re Looking For
People with post-production, footage-logging, media-cataloguing, captioning, subtitling, or audio-description experience who can describe footage accurately and consistently without embellishing or interpreting it.
Qualifications
- At least 1 year of relevant professional experience
- Bachelor’s degree, completed or in progress
- Frame-accurate attention to detail and fluency with timecode
- Comfort describing footage in a controlled vocabulary that is accurate, literal, and consistent
- Clean written English and solid transcription habits, including speaker attribution and non-speech sound labeling
- Must be based in US or Canada
Technical Requirements
- Desktop or laptop running Windows 10 (or later) or a recent version of macOS. Chromebooks, Linux-based systems, and tablets (e.g., iPad, Surface in tablet mode) are not supported.
- The self-direction to work independently on a flexible schedule
Bonus Points
- Stock or archival cataloguing and metadata tagging experience
- Audio-description experience describing visual detail precisely
- Familiarity with Avid, Premiere, iconik, CatDV, Axle, or similar media tools
What We Offer
- $28–30+/hour (USD) for productive time on the platform
- Paid three times a week through Stripe
- 100% remote and fully flexible, with 15+ hours per week recommended and no cap on hours
- A path to higher-paid Reviewer work for consistent, high-quality contributors
