$71M Backs Enigma's Hardware-Agnostic Robot Intelligence Layer

Enigma emerged from stealth with a $71 million seed funding round to build a unified intelligence layer for physical AI. The startup, founded by veterans of Israel’s Unit 8200, is developing hardware-agnostic foundation models designed to reduce the friction of programming robot behavior.

The Robots.online Data Collection

Coinciding with the funding, Enigma launched robots.online, opening a fleet of over 100 proprietary AI-powered robotic arms to remote public access. The fleet is distributed across physical hangars in California and Israel.

Users can manipulate the robots in real-time to perform illustrative tasks like painting, mixing chemical flasks, and fencing. The platform supports multiple control modalities, allowing users to direct the hardware through text, voice, video demonstration, or direct on-screen manipulation.

This large-scale public deployment serves as an active data collection engine. By observing how users naturally attempt to instruct physical machines, Enigma aims to optimize prompt engineering and interface design for complex physical tasks.

Hardware-Agnostic Foundation Models

The company’s core technical thesis treats instruction translation, rather than raw hardware capability, as the primary bottleneck in robotics. Enigma’s unified software translates natural language and direct inputs into physical actions across different robot architectures without requiring custom code for each deployment.

If you build multi-agent systems for industrial use, this approach attempts to abstract away the underlying kinematics. The startup developed both its initial models and the robotic hardware in-house to maintain tight control over the training loop. Enigma states its models require less manual data collection to maintain reliability in physical environments compared to traditional approaches.

Backing and Commercial Deployment

Index Ventures and Ribbit Capital co-led the seed round, with participation from Sarah Guo’s Conviction Partners. The cap table includes executives and researchers from major AI labs including OpenAI, Anthropic, DeepMind, xAI, Cognition, and Wiz. The founding team is led by CEO Jonathan Jacobi and CTO Gal Niv.

Enigma is currently testing its intelligence layer with early commercial partners in the healthcare, retail, and entertainment sectors to validate the software across distinct operational environments.

For teams deploying physical automation, track Enigma’s interface abstractions. As foundation models for robotics mature, the shift from explicit trajectory programming to intent-based UIs and hybrid agent interfaces will dictate how quickly these systems can scale in non-technical environments.