Sebastian Thrun's new company, Dulo, has raised $130 million in seed funding to build foundation models for hardware design, according to an exclusive from The Information on October 29. The investors named in follow-up coverage from Dealroom and TechStartups include Eric Schmidt, John Doerr, DST Global, Construct Capital and Lowercarbon Capital. No valuation has been reported, and the company has not put out a press release or a product page yet, so most of what we know comes from those reports and from Thrun's own brief comments over the summer.
For founders and engineers who work on physical products, this round is worth a close read. A seed check this large for a team still in stealth tells you how investors are pricing "AI for atoms" companies right now, and it points at a part of the stack (the slow, expensive work of designing and iterating on hardware) where a lot of Bay Area builders spend their days.
What Dulo says it is building
Dulo has kept its public description short. Reports describe it as building "foundation models for hardware design," and its web page, hosted on Stanford's robotics site, carries the tagline "manufacturing at lightspeed," according to AI Weekly. Pulse 2.0 reports that Thrun cofounded the company in 2026 with Adrian Aoun and that the stated goal is an "engineering superintelligence." Dealroom notes the product and customer plans remain unclear, which is normal for a company at this stage and a good reason to treat any detailed claims about its product with caution.
The team page, as reported by AI Weekly and The Next Web, says Dulo includes people from Waymo, Google Brain and Stanford's AI Lab. Thrun first named the company at the end of his keynote at Actuate, a robotics conference in San Francisco, on August 18. "I am not speaking about the company yet," he said at the time. "It's under stealth, it's very small. But it's in robotics."
Why Thrun's track record matters here
Thrun is one of the few people in Silicon Valley whose career runs straight through both AI research and physical machines. According to his Wikipedia biography, he directed the Stanford AI Lab, led development of Stanley, the Stanford robot car that won the 2005 DARPA Grand Challenge, cofounded Google X and led development of Google's self-driving car project, the work that became Waymo. He also cofounded Udacity, which was based in Mountain View.
That history explains why investors were willing to write a $130 million check before a product is public. It also hints at where the work is likely to happen. Stanford, Palo Alto and Mountain View are where much of the Peninsula's hardware and robotics talent already lives, and if you're building in that corridor, South Bay Renaissance, the in-person community Tribe organizes to rebuild the startup and social scene from San Mateo to San Jose, runs hardware and robotics gatherings for exactly this crowd.
Where Dulo fits in the AI-for-hardware wave
Dulo is arriving in a busy month for the category. On October 6, Palo Alto's Vinci raised a $250 million Series B at a $1.5 billion valuation led by Advent, Temasek and Xora Innovation for AI physics simulation in chip and hardware design, starting with heat. On October 28, London's Circuit Mind announced $25 million led by Norwest for tools that turn block diagrams into circuit schematics and check them with AI agents. Dealroom cites both as comparable deals.
These rounds sit inside a bigger shift in money toward physical AI. We covered the numbers in our breakdown of robotics startup funding in Q3 2026, when Crunchbase counted $9.9 billion in global robotics funding for the quarter, and in our look at Bedrock Robotics' $600 million round for autonomous construction equipment. The pattern across all of these is the same. Investors who spent 2 years funding software models are now funding models that touch the physical world, and they're paying up for teams with deep research and shipping experience.
What this means for hardware founders and engineers
If you run an early hardware company, the practical takeaway is to watch where design tools are heading over the next year. Simulation, schematic generation and design review are the steps that eat weeks of engineering time, and they're the first places AI tools from companies like Vinci and Circuit Mind are aimed. Founders who build habits around these tools early may be able to run more design iterations per dollar, which matters a lot when every board spin or prototype costs real money.
For engineers deciding where to work, a company with $130 million in the bank and a founder with Thrun's record will be hiring, and the reported mix of Waymo, Google Brain and Stanford alumni points to a high bar. It's also a reminder that the most interesting hardware jobs in the Bay Area right now often sit at the line between machine learning and mechanical or electrical engineering, so people with skills on both sides are in a strong position.
For investors, the round is another data point showing seed is now a loose label. A $130 million "seed" for a stealth company is closer to what a Series B looked like a few years ago, which resets expectations for every other physical AI team raising this winter, for better and for worse.
Meet the people building physical AI
The best way to track a category this young is to spend time with the people working in it. Our guide to robotics meetups in the Bay Area lists the recurring groups worth showing up to, and our own Physical AI Frontier session in downtown Palo Alto in October brought builders together to talk about moving AI from software into the real world.
If you want to keep the conversation going between events, the Futurists in Robotics Tribe is where members trade news like the Dulo round, share papers and compare notes on hardware tooling. For founder dinners, demo nights and robotics mixers around the Peninsula and South Bay, follow the Tribe Events calendar on Luma so new dates land in your inbox as soon as they're posted.
