Veeda AI

Veeda AI builds multimodal foundation world models that simulate physical reality. The product provides scalable virtual environments in which embodied agents learn by interacting with a simulated world.

The website argues that robots need to learn by trial and error in simulation, because real-world training is unsafe, expensive and slow. It does not name specific customers or describe a revenue model.

The company is headquartered in Toronto according to the registry, and the website lists locations in Toronto, Zurich, the Bay Area and Singapore. It describes itself as a small, deeply technical team and lists seven open roles, all in the Technical Staff family: world models, simulation, robotics, ML performance, ML data, ML operations and an internship.

Product: Multimodal foundation world models: scalable physical simulations in which embodied agents learn through interaction.