DIDERO.

The quiet engine behind the things you buy

Somewhere right now, a procurement manager is forwarding a PDF to herself. A supplier emailed a price change. A packing list says one quantity; the purchase order says another. Multiply that by four hundred orders, and you have a job. Didero wants that job.

Didero is a New York company that builds AI agents for procurement - the unglamorous machinery of buying parts, tracking orders, reconciling invoices, and keeping suppliers honest across time zones. Not a dashboard you log into and admire. A coworker that reads your inbox, understands what it means, and does the next thing. The company calls it "supply chain on autopilot," which sounds like marketing until you watch it close a purchase order loop without a human touching a key.

In February 2026 it raised a $30 million Series A. The round was co-led by Chemistry and Headline, with Microsoft's venture fund M12 along for the ride. That is real money for a company that, in the grand tradition of the most useful software, fixes a problem most people have never had to think about - and a handful of people think about constantly, badly, at midnight.

Global trade still runs on the inbox

The physical economy moves on procurement, and procurement moves on email. Spreadsheets. Phone calls. A messaging thread three apps deep. The big enterprise resource planning suites promised to fix this decades ago, and to their credit they did organize the data - they just never agreed to do the work. Somebody still has to read the supplier's reply, decide what it means, and update six systems by hand.

For mid-market manufacturers, the math is worse. They run complex international supply chains without the budget of a Fortune 100, and without the army of analysts to babysit the back office. The result is a profession that spends its days chasing exceptions instead of making decisions - the corporate equivalent of bailing a boat with a teaspoon while being told to plot the course.

The insight Didero bet on is almost rude in its simplicity: most procurement work is language. Read this email. Compare it to that order. Notice the lead time slipped. Send the follow-up. Language is precisely the thing large AI models got good at. So the question was never whether software could understand a packing list. It was whether anyone would trust it to act.

Three people who had felt the pain

Co-founder and CEO Tim Spencer did not discover this problem in a market-sizing slide. He lived it. During the pandemic he ran an e-commerce startup, Markai, where his team drowned in manual procurement just trying to keep shelves stocked. He sold that company in 2023 and, instead of resting, went looking for the part that had hurt the most. He found Didero, alongside co-founders Lorenz Pallhuber and Tom Petit, in late 2023.

Their bet was a particular one. Not "build a smarter dashboard," but treat AI as an autonomous operator - something that adapts to changing conditions and makes calls based on real-time data about pricing, lead times, and quality issues. The agent sits as a layer on top of the systems a company already runs, rather than asking anyone to rip out their ERP. In a category famous for eighteen-month implementations, "no IT build work" is its own kind of radical.

An agent that reads the mess, then acts

Didero's procurement agent ingests the natural-language sprawl of a real supply chain - emails, messaging apps, phone calls, purchase orders, packing lists - and turns it into executed work. It can source and quote, manage the full purchase-order lifecycle, process invoices, and handle the back-and-forth with suppliers. Then it writes the result back into the tools the team already uses.

The numbers behind the bet

Conviction in this category is measured in two currencies: who writes the checks, and who runs the software in production. Didero has collected both. The funding tells you investors believe the back office of global trade is up for grabs. The customer count tells you operators agree.

The customer list reads like the supply side of everyday life - the people who actually make and move things.

An autonomous supply chain, on purpose

Didero's stated aim is to build an autonomous supply chain - one where the routine, manual work runs itself so people can spend their hours on judgment instead of data entry. It is a mission that benefits from being measurable. Either the agent closed the loop or it did not. Either the team got its afternoon back or it is still forwarding PDFs.

The company is careful about the framing. The agent is a coworker, not a replacement - it takes the work nobody wanted, the chasing and the copying and the reconciling, and hands back the part humans are actually good at: deciding. That is a quieter promise than most AI companies make, which is probably why it lands.

The boring layer is the strategic one

Tariffs shift. Lead times stretch. A supplier two continents away misses a date and a factory here goes quiet. The companies that handle disruption best are not the ones with the prettiest dashboards - they are the ones that can sense and respond fastest, across thousands of small decisions a day. That capacity is exactly what an agent that lives in the operational layer can provide.

With a Series A behind it and the Covalent AI acquisition folded in, Didero is pushing its agent further across sourcing and payments. The wager is that procurement stops being a cost center to be tolerated and becomes a system that quietly compounds - resilience as a feature, not a fire drill.

So picture that procurement manager again. It is 7am. The inbox already holds two hundred emails - the price change, the mismatched packing list, the supplier who went quiet. Only now most of them are already handled. The price change got logged, the mismatch got flagged, the follow-up got sent at 3am by something that does not sleep. She opens her laptop not to bail the boat, but to steer it. That is the change Didero is selling, and the part it would rather you judge for yourself.