All jobs
Rilith logo

AI Data Collector (Remote)

Rilith · Posted today

  • India (Remote)
  • Remote
  • Part-time

About the role

Role: AI Data Collector — Egocentric Video Recording

Location: Remote — open to all countries worldwide (excluding China)

Job Type: Part-Time / Flexible

Commitment: Flexible — fit around your existing routine

Language: No language requirement — task recording only, no speech needed

Experience: None required — everyone is welcome

About the Role

Rilith is partnering with students, working professionals, and homemakers to build one of the world's most comprehensive first-person video datasets for AI and robotics research

.Your role is simple: record short, first-person (egocentric) videos of everyday tasks at your home or workplace — cooking, cleaning, laundry, warehouse operations, retail activities, and more. No acting, no scripts, no extra time. Just your normal routine, captured through a head-mounted phone strap. Your contribution directly trains the intelligent systems of tomorrow.

Eligibility

  • Every model from iPhone 12 to iPhone 18 (all variants)

Hardware

You will need a mobile head-mount strap (cost min Rs. 300)— a lightweight rig that holds your phone hands-free in first-person position

  • Reimbursement — complete the project, and your strap cost is fully covered
  • Recover the cost fast — even 1 hour of recording covers what you spent
  • Use it beyond this project — Rilith runs similar projects regularly

Payment

  • Pay rate: Rs. 250 per hour of approved recording
  • Payment frequency: Weekly payment
  • Why Join?
  • Contribute to next-generation AI and robotics — your footage has real, lasting impact
  • Record during your normal daily routine — zero extra time commitment
  • Students and professionals can request an official Certificate of Contribution on completion
  • Your data stays secure — nothing is shared without your explicit consent
  • 24/7 expert support throughout your participation
AI Data Collector (Remote) at Rilith, India (Remote) | Quark9