Training data made by real people, checked twice.
We collect and label video, image and audio data for AI teams, with a screened workforce across 55 countries and two levels of QA on every delivery.
Everything between raw data and a clean dataset.
Scoped to your guidelines, priced per accepted unit, delivered in your format.
Data collection
Egocentric and workplace video, photos, speech and voice, recorded with signed consent from every participant.
Annotation
Boxes, polygons, keypoints, tracking and action timelines, on our tools or inside yours.
Audio & language
Transcription, speaker labels, translation and voice-over narration in many languages.
Managed teams
Screened annotators on your platform, with team leads, shift cover and daily reports.
From brief to delivery in four steps.
- 1
Scope
Guidelines, volume and acceptance criteria agreed in a short SOW.
- 2
Screen
Every contributor passes a test built from your guidelines.
- 3
Pilot
A small paid batch you review before we scale up.
- 4
Deliver
Two-level QA, then files in your format and naming convention.
A screened workforce in 55 countries.
Recruit by country, language, age group or skill. Our largest teams are in the Philippines and India.
Every recording comes with signed consent from the people in it.
We never resell client data, and we delete it when you ask.
Client NDAs, plus confidentiality agreements with every contributor.
We follow the Philippine Data Privacy Act (RA 10173). Privacy Policy
Have a dataset in mind?
Send your brief below. We reply with a screened team and a pilot plan.
Send a briefEarn from home on AI projects.
Record, label and transcribe on flexible hours. Paid per approved task. We never charge any fees to join.
See open projects