Enigma raises $71M to develop foundation models for robots

UPDATED 19:18 EDT / JULY 27 2026 AI Enigma raises $71M to develop foundation models for robots by Maria Deutscher Engima Ltd., a provider of artificial intelligence software for robots, launched today with $71 million in funding. Index Ventures and Ribbit Capital jointly led the seed round with participation from Conviction Partners. They were joined by a group of angel investors that included employees at Google DeepMind, Anthropic PBC and OpenAI Group PBC. Teaching an industrial robot a new task usually requires developers to write custom code. Furthermore, the code has to be adapted to every environment in which the machine operates. Two identical robotic arms installed next to differently-sized conveyor belts require two different configuration scripts. Equipping a robot with AI eases the programming workflow. Some neural networks, such as Nvidia Corp.’s GR00T, make it possible to configure their host machines with natural language prompts. That saves a significant amount of time for engineers.  San Francisco-based Enigma is developing foundation models designed to “make robots intelligent and effortless to use.” The company was founded last year by cybersecurity researchers Jonathan Jacobi (pictured, right) and Gal Niv (left). Jacobi, Enigma’s Chief Executive Officer, joined Microsoft Corp. at age 17 as its youngest-ever employee. He later worked at a startup that was acquired by Google LLC’s Wiz cybersecurity business. Enigma says that its models are capable of running on practically any robot. As a result, customers can adapt the models to a new machine with a fraction of the training data usually required for the task. That addresses one of the main challenges involved in developing AI-powered robots. Developers usually teach an AI model how to steer a robot by showing it footage of the robot performing various tasks. Such footage is more expensive to collect than many other types of training data. In some cases, it requires companies to create mock production lines and record how robots interact with them. Reducing the amount of training data required for robotics initiative lowers deployment costs and saves time. Enigma is not the only company working to shrink robots’ training datasets. Last year, Meta Platforms Inc. released a model called V-JEPA that can be adapted to a new robot with under 100 hours of explanatory footage. The key to the model’s adaptability is a neural network architecture that Meta introduced in 2022. It enables AI models to predict future events in a factory and optimize their decision-making accordingly. Besides AI models, Enigma is also developing an interface for controlling robots. The company today launched a website that enables users to interact with more than 100 robots optimized for various tasks. Enigma will use the data generated by those interactions to refine its interface.  “We believe AI’s next chapter is moving beyond chatbots and screens into systems that can understand, adapt to, and operate in the physical world,” Jacobi said. “We’re building the intelligence layer for robotics through breakthrough AI models and novel human-robot interfaces.” Enigma will use its seed funding to hire more engineers and purchase computing capacity.  Photo: Enigma A message from John Furrier, co-founder of SiliconANGLE: Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities. 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders s