training data infrastructure for robotics
mission

We are building the largest provider of high-quality, high-volume training data for large-scale robotic models — treating data production as industrial infrastructure, not a research problem.

approach

Data is the resource the world runs on. In robotics, that moment is arriving now — and the data problem is orders of magnitude harder, and we believe larger.

Every major technological shift has been a data shift. Robotics is next — and the scale of the data problem dwarfs anything that came before it. We collect the data that actually matters. The kind that's hard to get — in real environments, under real conditions, with operators who know what they're doing. Every capture is clean. Every edge case is documented. Every dataset ships verified, versioned, and ready to train on.


team
Rohan Seelamsetty Cognitive Science, Penn · Apollo Fellow, Oxford
Dev Karpe Huntsman Program, Penn · Apollo Fellow, Oxford
Ishan Ahluwalia AI + Applied Math, Penn · GRASP Lab

contact

ishan@praxisrobotics.io