Enigma emerged from stealth on Monday with a $70 million seed round led by Index Ventures and Ribbit Capital, alongside Sarah Guo of Conviction Partners. The less-than-year-old research lab is betting that the bottleneck in robotics is not model capability but the interface between humans and machines — and it plans to test that thesis by letting anyone on the internet control more than 100 of its proprietary AI robots.
The robots sit in hangars in Israel and California. Users can direct them to draw pictures with a paintbrush, fence with swords, or run simple chemistry experiments by picking up and mixing flasks. Enigma built both the robotic arms and the underlying models from scratch, an unusually vertical footprint for a company at seed stage. The $70M is the raw material for a data-collection experiment rather than a product launch.
The founders are outsiders to robotics. Jonathan Jacobi was Microsoft's youngest-ever employee, recruited during his stint there by Wiz founder Asaf Rappaport. He co-founded Enigma with Gal Niv, a friend since teenage hacking competitions and later a colleague in Israel's Unit 8200 cybersecurity corps. Their founding team is drawn from Israel's tech ecosystem — alumni from AI labs, math Olympiad winners, and several people who dropped out of PhD programs to join.
Key facts
- 01Enigma raised a $70M seed round led by Index Ventures and Ribbit Capital, with Sarah Guo of Conviction Partners participating.
- 02The lab is opening 100+ proprietary AI robots — housed in Israel and California — to online control by anyone in the world.
- 03The company is less than a year old and built both the robotic arms and their underlying models from the ground up.
- 04Co-founders Jonathan Jacobi and Gal Niv met in Israel's Unit 8200; Jacobi was previously Microsoft's youngest-ever employee, recruited by Wiz founder Asaf Rappaport.
- 05Enigma says it is already partnering with companies in healthcare, logistics, and entertainment, though it declined to name use cases.
That outsider status is part of the investment thesis. Index Ventures' Shah frames Enigma's angle as an experience question rather than a capability question, in contrast to teleoperation- or dexterity-first competitors. The pitch is that fresh eyes on human-robot interaction may find a shape that roboticists trained inside the field would not think to look for.
Jacobi's operating analogy is a car's volume knob. Users would revolt, he argues, if adjusting audio required specifying percentages blind. The same friction, he says, describes the current state of general-purpose robots — including ones running the most capable available models.
The public experiment is designed to surface which input modes actually work: text, voice, video demonstration, or direct tap-drag-drop manipulation of what the robot sees. Enigma expects the data to inform not just the interface but the training pipeline for its foundation model. If the answer to 'how should humans talk to robots' is different from 'how should robots talk to each other,' the model architecture may need to reflect that.
The approach cuts against the two dominant playbooks in the current robotics wave. One camp is training on millions of scraped web videos to give models a passive prior on how the physical world behaves. Another is collecting first-person data from humans wearing sensorized gloves. Enigma is generating its data by putting strangers in the driver's seat of real hardware and watching what they instinctively try.
What Enigma will sell is still open. Jacobi declined to specify use cases but said the company is already working with partners in healthcare, logistics, and entertainment — three verticals with wildly different tolerance for latency, error, and cost. A $70M seed is large enough to sustain that ambiguity for a while, and the investor syndicate signals patience for a research-heavy roadmap.
The risk is legible. Open-ended research bets in robotics have a long history of producing interesting demos and thin revenue, and "we'll figure out the interface" is a harder story to underwrite than "we ship warehouse pick-and-place." Enigma is also entering a crowded field where well-funded rivals — from humanoid startups to embodied-AI labs inside the frontier model providers — have head starts on dexterity and on paying customers. If the public experiment produces novelty but no repeatable interaction pattern, the company will need a second act.
The interesting wager here is the sequencing. Most robotics companies are building capability first and hoping usability follows; Enigma is building usability first and hoping capability follows the data. That inversion is only affordable because the seed is large enough to fund the hardware fleet and the model work in parallel. If the public experiment surfaces a genuine interaction primitive — the volume-knob equivalent — the payoff is a defensible layer sitting between every future humanoid and its user. If it doesn't, Enigma will have spent $70M running the most expensive UX study in robotics history.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
