He teaches software to read a supplier's frustrated email at 2am, find the late shipment, and fix the purchase order before anyone walks in for coffee.
CO-FOUNDER & CEO → DIDERO · NEW YORK
The quiet engineer behind a noisy idea
Tom Petit spends his days on a problem most people are happy to ignore: the supplier message that says a part is three weeks late, buried in an inbox, discovered on the day the line was supposed to run. His company, Didero, is an AI procurement agent that lives on top of a manufacturer's existing systems, reads the incoming flood of supplier emails and documents, and quietly does the routine work of keeping a supply chain moving. Purchase orders, supplier onboarding, invoice reconciliation, the dull connective tissue of global trade. The agent handles it.
The pitch is unfashionably specific. Didero is not trying to reinvent the enterprise resource planning systems that big manufacturers already run. It sits above them, a coordination layer, parsing what humans currently parse by hand. Petit calls the goal "the autonomous supply chain, where all the routine and manual work can be abstracted away." In February 2026 investors agreed enough to write a $30 million Series A check, co-led by Chemistry and Headline, with Microsoft's venture fund M12 along for the ride. That brought Didero's total funding to roughly $37 million.
What makes the bet interesting is who is making it. Petit is a computational mathematician by training and a data scientist by trade, the kind of person who did analytical work at Airbnb and HelloFresh before deciding that supplier coordination was the problem worth a decade. He is also a second-time founder, which means he has already learned what most first-timers spend years discovering.
"$32T of international trade binds nations together, yet supply chain teams are not given the tools they need."
He grew up in Belgium and moved to Texas, and he talks about the jump as the thing that rewired him. Texas felt bigger, more ambitious, more bold than the careful Europe he came from. You can hear both halves in how Didero is built: European exactness about the mechanics of a purchase order, American appetite for a market worth trillions.
The resume reads like a deliberate education in how systems break. Data science at Airbnb, where supply and demand never quite line up. Data science at HelloFresh, where a single late ingredient ruins a hundred thousand dinners. Then Landis, the Sequoia-backed startup he co-founded to help renters become homeowners, a fintech tangle of underwriting and trust. Each one was a lesson in coordinating messy human processes at scale.
Between Landis and Didero, Petit did something founders rarely admit to: he slowed down. Rather than sprint into the next idea, he took time to reconnect with why he builds at all, letting curiosity steer instead of momentum. The honesty extends to how he talks about the job. He once assumed successful founders had godlike domain expertise. Then he became one and concluded that "most companies aren't really rocket science" and that founders, talented as they are, aren't a different species.
Didero itself was born from a co-founder's scar tissue. Tim Spencer had run an Asia-based e-commerce company, Markai, where keeping thousands of suppliers across dozens of countries in sync meant endless manual coordination. Lorenz Pallhuber brought years inside McKinsey's procurement practice. Petit brought the machine learning. Three people, one shared conviction that the boring middle of global trade was quietly drowning in email.
Purchase orders and supplier emails are nobody's dream startup. That is exactly why Petit chose them. The unglamorous middle of trade is where the money and the misery actually live.
Computational mathematics at Stanford and Cambridge. He aimed an elite quantitative education at one of the least quantitative-looking corners of business.
He started out believing the people who build companies were a breed apart. Building one taught him they aren't. It made him calmer about the job, not less ambitious.
"We're not creating our own foundational models," he says, "but we're doing a lot of fine-tuning." Pragmatism over ego. Use what works, sharpen it for the job.
Beyond building, he has backed more than 15 startups. The pattern: people who build the unsexy infrastructure other people depend on.
Ask him about New York in July and you'll get a tip, not a brand: "Governor's Island is such a hidden gem." Small detail, real person.
"Our objective is to build the autonomous supply chain, where all the routine and manual work can be abstracted away."
"We're using a lot of foundational models and APIs that are out there. We're not creating our own, but we're doing a lot of fine-tuning."
"$32T of international trade binds nations together, yet supply chain teams are not given the tools they need."
"Governor's Island is such a hidden gem."
Didero in 2025 was a roughly 14-person team reportedly running near $1.5 million in revenue, the awkward in-between stage where a product works but the market is still being convinced. The Series A is the bet that the next chapter is bigger: more manufacturers, more of the procurement workflow handed to agents, fewer 2am emails discovered too late.
The ambition is plain in how Petit frames it. Not a feature, not a dashboard, but the autonomous supply chain itself, the routine coordination of global trade running on its own so the humans can spend their hours on judgment instead of busywork. It is a large claim from a measured person. The measured ones are sometimes the ones to watch, because they tend to know exactly how hard the thing they promised actually is.