Real-world training data, collected to your brief
You write the spec. Consenting adults record it on their phones or head-mounted cameras, and a person checks every clip against your spec before it counts. You get a quote by email before anything starts.
Who we collect for
- Robotics and physical AI teams who need people doing real tasks in real homes.
- ML teams who need audio, images or handwriting a public dataset does not cover.
- Agencies and data vendors who need a collection partner for a client brief.
What we collect
- First-person video of household tasks, filmed at head height with the hands and objects in view
- Long-form point-of-view recording of everyday life
- Everyday objects filmed from set angles
- Ambient household sound
- Handwriting samples written to your phrases
- In-person sessions with devices such as smart glasses
What you control
- The capture instructions
- Video or still images
- Minimum and maximum clip length
- Whether audio is required
- File formats and maximum file size
- How many submissions you need
- Which countries contributors can be in
Every clip is checked before it counts
A person reviews each submission against your spec.
- A clip that misses the brief is rejected with a reason, and the contributor can record it again while the task is open.
- Contributors are paid only for accepted work, so they are paid to follow your brief.
Consent and licence you can show your legal team
- Contributors are adults who confirm they are 18 or over and only submit media they are allowed to share.
- The consent terms are recorded with each task.
- Contributors grant a licence to train and evaluate machine learning models, including on behalf of GoKuna's customers.
What each collection looks like
| Collection | What contributors do | Contributor pay | Time |
|---|---|---|---|
| Object videos | One ordinary object on a clear surface. A steady 10–20 second clip showing the whole thing from two sides. | $5.00 per accepted clip | about 3 minutes |
| Audio | A few seconds each of a kettle, a tap and a drawer, recorded where you normally cook. No voices. | $6.00 per accepted set | about 4 minutes |
| Document photos | Two short phrases written in your own hand, photographed flat in good light. Nothing personal in the words. | $8.00 per accepted sample | about 5 minutes |
| First-person video | Short point-of-view clips of ordinary jobs around the home, such as cooking, tidying and laundry, filmed with a phone or action camera mounted at head height. | $10.00 per accepted clip | Worldwide |
| Long-form video | Longer stretches of everyday life recorded from a head or chest mount, showing normal activity as it happens rather than staged scenes. | $20.00 per accepted hour | North America to start |
| In-person sessions | In-person sessions trying out a smart-glasses prototype built to help with hearing loss, and describing how well it works. | $50.00 per session | New York area |
What happens after you send a brief
- We reply by email with a quote and any questions about the spec.
- We write your spec as the task contributors will read, and you check the wording.
- A small first batch is recorded and reviewed against the spec, so you can check it in your own pipeline.
- Once the spec is right, the full volume runs on the same brief.