United States
Role: AI Data Collector (Remote)
Location: Work from anywhere
Job Type: Part-Time / Flexible
Daily Commitment: Flexible
No language barriers — this project records tasks only; no audio or speech required.
No experience required.
About the Role
Rilith AI is partnering with contributors — students, working professionals (office/field), and homemakers — to work on a research project in which participants record short egocentric (first-person POV) videos of everyday household tasks and daily activities performed at their work locations. You will record your hands carrying out real-life activities — at home or at work — with no extra time required beyond your normal routine. Everybody can participate.
Each participant can record up to 100 hours per environment (e.g. home, warehouse).
Eligibility Requirements
You must:
What You Will Do
Record first-person (egocentric) videos while performing real-life activities, including:
Check the app for the full list of available tasks.
Bonus & Payment Terms
Why Join?
Your data will remain secure. No personal information will be shared or disclosed without your consent.
About Rilith:
Rilith is empowering intelligent machines at scale. With teams spanning the USA, LATAM, EU, and Asia, we bring a truly global perspective to every project we undertake. Our on-the-ground experience extends beyond software and data — our team has hands-on experience working directly alongside robotic hardware, giving us a deeper understanding of real-world deployment needs. At Rilith, we take data responsibility seriously; all footage we collect is PII-free, with no face capture, ensuring every deliverable is fully privacy-safe. We don't just collect data — we collect it the right way.
We are highly agile and adaptive to challenging data projects.
Specializations:
Egocentric Head-Mounted Video – First-person POV captures using head-mounted rigs across residential and commercial settings. We handle participant recruitment, field deployments, and high-quality footage delivery.
Egocentric Human Hand Video – Hand-object interaction recordings in manufacturing, household, and daily-use settings using a multi-view rig (head, left-wrist, and right-wrist cameras).
Robotics Manipulation – Video capture of robotic gripper (2–3 finger) and dexterous hand (4–5 finger) tasks using a multi-view rig (head/overhead, left/right gripper, and front/top views).
Join our global network of Rilith Warriors:
Rilith has expertise in creating data specifically for Physical AI, which it does in collaboration with a globally spread freelancer network — we call them Rilith Warriors. What differentiates us from others is that we respect, care for, and understand the challenges faced by freelancers.
We ensure: