Train Them AI is a data ops boutique for Seed–Series B robotics teams that need policy-ready VLA datasets without the $500K minimums and 6-month timelines of enterprise vendors.

We do three things:

• Egocentric data collection — distributed operator network across Latin America and the Philippines, capturing first-person video of household, kitchen, and workspace activities at scale.

• Teleoperation data collection — portable Quest 3 + GELLO setups, on-site or remote, on your hardware (ALOHA, UR, Franka, custom rigs).

• VLA action annotation — episode-level language instructions, subtask segmentation, success/failure labels, contact tags, trajectory quality scores. OpenVLA · pi0 · Octo · ACT.

Every delivery includes a policy smoke test before we ship. We run a mini training pass on 10% of the dataset and fix the data if the policy doesn't converge — no other vendor in our segment does this.

Pilots start at 100 demos and ship in 2 weeks, delivered in LeRobot v2/v3, RLDS, HDF5, or robomimic. No hardware minimums. No enterprise sales cycle.

Built for the teams that raised $2M–$50M, have robots in the lab, and are hitting the data collection bottleneck — but can't get Scale AI's attention.

trainthemai.com

Train Them AI

Org chart

Teams

Offices