Timmy Seto Turns Machine Shops Into Software
Most engineers wait weeks for a single metal part. The co-founder of Forge Automation decided that number should be four days, and built a company around making it true.
Ask most founders what they do and you get a paragraph. Ask Timmy Seto and you get a sentence: he makes metal parts. Send him a CAD file, and he will make it exist in the real world. That plainness is not a lack of ambition. It is the whole point. Behind it sits Forge Automation, the Toronto company he co-founded and runs, which is trying to do something the hardware world has quietly wanted for years - make a custom metal part show up in four days instead of three weeks.
Seto is the co-founder and CEO of Forge Automation, part of Y Combinator's Winter 2025 batch. The pitch is simple enough to fit on a napkin. An engineer uploads a CAD model to a website, and a software-enabled factory turns it into a finished CNC part and ships it, fast. The customers are the people building physical things that are hard to build: robotics teams, consumer hardware startups, hard tech companies that live and die by how quickly they can iterate.
A decade at the machine
The confidence in that line is earned. Seto says he has been machining for more than ten years, which means he was standing at machines well before he was old enough to legally sign a company into existence. He went on to study engineering at the University of Waterloo, one of the few programs where students spend as much time inside real companies as inside lecture halls.
He used that time well. His co-op internships read like a tour of the places that care most about precision: Apple, Formlabs, and iRobot. Each one is a company where a physical product only works if the parts behind it are made exactly right. Seto watched how those teams operated, and he noticed something that would later become a business - even the best hardware companies get stuck waiting on parts.
The bottleneck nobody talks about
In hard tech, the exciting story is always the idea - the robot, the device, the machine that does something new. The unglamorous truth is that progress often stalls somewhere much more boring: a mechanical engineer needs one custom bracket, sends the drawing to a machine shop, and then waits. Two weeks. Sometimes three. During that wait, the whole project sits still.
Seto's insight was to treat that wait as the product. If you can compress lead time, you are not just saving days. You are changing how fast a hardware company can think. A team that gets parts in four days can try five versions in the time it used to take to try one.
Two engineers, no gap year
Seto did not build Forge alone. He co-founded it with Walter Raftus, now the company's CTO, whom he met through Waterloo engineering. Raftus came at hardware from the software side - AI and mechatronics, with internships at Nvidia, Intel, and Trexo Robotics. Where Seto knows the feel of a cutting tool, Raftus knows the code that can coordinate a floor full of them. The pairing is the company in miniature: metal and software, working the same problem.
Both graduated in the summer of 2022. Most of their classmates took full-time jobs at the kinds of companies Seto had just interned for. The two of them started a factory instead. There was no gap year, no long deliberation. They walked out of school and began making parts.
How the four days actually work
The magic is not a single machine. It is the software wrapped around a real factory floor - the layer that takes a CAD upload, turns it into a quote, schedules the work, and keeps the promise of a delivery date. Forge even offers a program where customers reserve dedicated machine capacity with a guaranteed turnaround, which is closer to buying a subscription to speed than to placing a one-off order.
It is worth pausing on how counterintuitive Seto's career move was. He had a clear path into any of the well-known hardware companies he interned for. He chose instead to build the supplier that sits underneath all of them. It is a less obvious kind of ambition - not making the shiny product, but fixing the thing every shiny product secretly depends on.
The quiet layer
There is a version of the startup world that is loud - the launches, the demos, the crowded timelines. Seto is working a quieter layer of it. Roughly two dozen people in Toronto, machines running, parts going out the door. It is the kind of business that does not trend, but that a growing wave of robotics and hardware companies increasingly cannot function without.
His ambition, stated plainly like everything else, is to make rapid, software-driven manufacturing the default rather than the exception - the Amazon experience for metal parts, where ordering a precision component feels as ordinary as ordering anything else online. If robotics keeps booming the way it has, the companies riding that wave will all need the same thing: parts, fast. Seto positioned himself exactly where that demand meets its oldest bottleneck.
For now he keeps describing the work the same way he always has. Send him CAD. He will make it real. It is a strange thing for a founder to make so little of, until you realize how few of them can actually do it.