Sebastian Thrun has raised $130 million in seed funding for Dulo, a new startup that builds AI models to automate engineering design. The Information first reported the round on October 29. Thrun founded Google X, and his Stanford team’s work became the starting point for Waymo. Dulo’s investors include former Google CEO Eric Schmidt, Kleiner Perkins chairman John Doerr, DST Global, Construct Capital and Lowercarbon Capital.
That’s a huge check for a company that, until this week, had no disclosed funding, no product and no timeline.
🧭 The Quick Read
- What: Dulo is building “foundation models for hardware design.” Foundation models are large, general-purpose AI systems, and these ones are meant for physical products instead of text or code.
- Who: Sebastian Thrun and co-founder Adrian Aoun, an ex-Google executive who founded the healthcare startup Forward. The team includes people from Waymo, Google Brain and Stanford’s AI lab (SAIL).
- How much: A $130 million seed round, very large for a first raise.
- Where: San Francisco. The company was founded earlier in 2026 and first came to light in August.
- The goal: Aoun says the mission is to build an “Engineering Superintelligence.”
⚙️ What Dulo Is Actually Chasing
Most of the AI money over the last three years has gone into language, images and software. Coding assistants turned into a business worth billions because code is text, and text is what large models handle best.
Hardware is a tougher problem. Designing a motor mount, a drone frame or a heat exchanger means dealing with geometry, materials, tolerances, physics and manufacturing limits all at once. Engineers still do much of this by hand in CAD (computer-aided design) software and run simulations to check their choices. It’s slow, it costs a lot, and progress depends on how many skilled engineers you can hire.
Dulo is betting that one model trained across all of those signals could propose, test and refine designs the way coding models now write and debug software. If that works, AI moves out of the IDE and into the factory.
📈 Why This Matters
What stands out is the size of the round compared with how little Dulo has shown. A $130 million seed is the kind of bet investors make on people, and Thrun has a strong record of turning research into new industries:
- Self-driving cars: His Stanford team won the 2005 DARPA Grand Challenge, and that work fed directly into Google’s self-driving project.
- Google X: He founded the moonshot lab that produced Waymo and other projects.
- Udacity: He co-founded one of the first big online education platforms.
- Kitty Hawk: He led the flying-car startup backed by Larry Page.
The round also fits a wider trend. Investors are moving money from chatbots toward “physical AI”: robotics, manufacturing, chip design and anything else that touches the real world. Huge seed rounds for star founders have become normal in 2026. Dulo now joins a group of labs that raised nine figures before releasing anything.
⚠️ The Open Questions
Some caution is warranted. Dulo hasn’t said publicly what it will ship first, who its customers are, or how it will get training data. That last one is the hard part. The internet is full of text and code, but good engineering design data mostly sits inside private company files.
Turning a model into a product will also be tough. Engineering teams work inside established tools like CAD suites and simulation software, and a bad design suggestion can cost far more than a buggy line of code. A big funding round doesn’t show that the approach works. It just gives Dulo time to try.
🔭 What to Watch
If you work in hardware, manufacturing or industrial design, pay attention to three signals:
- First product scope: Does Dulo start narrow (a single part category or simulation task) or go straight for general-purpose design?
- Data partnerships: Deals with manufacturers or CAD vendors would show how Dulo plans to solve its data problem.
- Hiring: Watch whether it hires mechanical engineers or mostly ML researchers. That will show what kind of company it plans to be.
AI coding tools changed software engineering in about two years. Dulo is wagering that hardware is next, and with Thrun leading it, many people will be watching. You’ll find the full report at The Information.