General Robotics

General Robotics is launching Auto-Engineering on GRID, its robot intelligence platform, extending the system so it can take a task description and automatically handle the integration, simulation, training, testing, and hardware validation work that normally requires a team of specialized robotics engineers. The launch video demonstrates four "robotics harnesses" that map a robot, build a simulated world with success criteria, engineer the skill itself, and then deploy and evaluate on hardware, with results flowing back into GRID so every deployment improves the next one. In the demo, the same pouring task is transferred from a Flexiv arm to a UR5E, extended to stirring, then to swirling a flask trained from a single iPhone video via video-to-simulation, and finally to a reactive policy that tracks a moving beaker. The product is aimed at industrial operators and robot makers trying to move past pilots into production. GRID customers span industry leaders across automotive manufacturing, port operations, energy generation, and food and beverage production, as well as multiple government agencies, and the platform is now integrated with robot manufacturers including Fanuc and Galaxea. The pitch this week is concrete numbers on that fragmentation problem, with the company reporting robot onboarding cut from roughly a month to as little as two hours, model ingestion from three days to twenty minutes, and skill transfer across form factors from three days to about an hour and a half. General Robotics is a Redmond, Washington startup founded in 2023 by Ashish Kapoor and Dinesh Narayanan to build composable AI skills that work across robot form factors, with co-founders Sai Vemprala (CTO) and Shuhang Chen coming from the same Microsoft research team . Investors include Construct Capital, Accenture Ventures, E14, NVIDIA, Shorooq, Valo Ventures and Khosla Ventures.