Mowito raises $3 million to scale physical AI for robots

Mowito, a startup building physical AI foundation models for industrial robot arms, raised $3 million in a pre-seed round led by Version One Ventures, Business Standard and Economic Times reported on July 7, 2026. All In Capital, Unisol, iSeed, and angel investors including Soumith Chintala, Adarsh Kulkarni, Ashish Kulkarni, and Vaibhav Domkundwar also participated. Founded in 2024 by Puru Rastogi, Adityanag Nagesh, and Safar V, the company operates teams in Bengaluru and Detroit and says the funding will accelerate its U.S. expansion, grow its engineering and go-to-market teams, and scale deployments across automotive and electronics manufacturers. "Manufacturing has reached a point where hardware is no longer the bottleneck, software is. Factory robots shouldn't need to be reprogrammed every time production changes. We believe robots should learn the same way people do: by observing and repeating," said CEO Puru Rastogi.

Mowito's pitch is that its software runs on standard, unmodified industrial robotic arms and teaches them new tasks from human demonstration rather than hand-coded motion scripting, an approach commonly called imitation or demonstration learning. That typically combines visual perception for pose and object-state estimation, models that convert observed demonstrations into robot-executable trajectories, and closed-loop control to hit manufacturing tolerances. The company says its robots are already deployed at a Fortune 500 automotive company and one of the world's largest electronics contract manufacturers, though neither customer is named and no cycle-time, defect-rate, or accuracy figures were disclosed.

If demonstration-based teaching genuinely works on unmodified hardware at manufacturing tolerances, the practical payoff is faster changeover when a production line's tasks change, without new control code for every SKU or process tweak. The open questions that determine whether that holds up in practice are calibration and repeatability across different arm models, safety certification for closed-loop execution near human workers, and how reliably the approach transfers from a demonstrated task to novel variations on the line.

All In Capital's Kushal Bhagia called Mowito's technical depth and early customer validation central to the investment thesis; watch for named customers, measured production outcomes (cycle time, defect rate, uptime), and whether the company discloses technical detail on calibration and safety as it scales in the U.S.