Here is a fact about big companies that sounds made up but is not: procurement - the unglamorous business of buying the things a company needs, from laptops to legal software to industrial lubricant - is usually the second-largest thing a company spends money on, right after paying its own people.
And here is the second fact, which is the whole reason Levelpath exists: the software that manages all that money is, more or less universally, despised by the people forced to use it. This is a strange state of affairs. Companies pour enormous sums through these systems and then treat the systems themselves like a trip to the DMV - something you route around, delay, or hand to an intern. Levelpath, a San Francisco company founded in 2022, looked at that gap between how-much-money and how-bad-the-software and decided it was a business.
The pitch, stripped of jargon, is roughly: what if procurement software were good? Not "good for procurement software," which is a famously low bar, but good in the way a well-made consumer app is good - fast, mobile, obvious, the kind of thing you don't have to be trained on for three days. Levelpath calls its platform "AI-native," which is a phrase that has been doing a lot of heavy lifting in enterprise software lately. In their case it means the artificial intelligence was not bolted on as a feature after the fact but built in as the substrate - agents that read messy contracts, draft sourcing events, analyze supplier bids, and generally do the tedious parts that a person would rather not.
The founders have done this before
You are allowed to be skeptical of a startup promising to fix an old, boring, entrenched category. Many have tried. What makes Levelpath worth a second look is that its founders already did roughly this, and it worked. Alex Yakubovich (the CEO) and Stan Garber are high-school friends from Ohio who have been building things together for more than twenty years. Their previous company, Scout RFP, was a procurement platform that Workday acquired in 2019 for $540 million. They then spent a stretch inside Workday watching, up close, all the ways procurement stays broken even after a big acquisition - and, apparently, could not help themselves. They left and started over, this time with two additional co-founders, Bryan Rosenstein and Raimonds Samofals.
There is a tidy way to read this: the $540 million exit was not the finish line, it was the research phase. The founders learned exactly where the old model fell short and built the next one to attack those specific gaps - mobile-first, AI-native, and obsessive about whether people actually like using the thing.
The money, and who is behind it
Levelpath has raised in three acts. A $14.5 million seed round led by Benchmark got it off the ground. A $30 million Series A led by Redpoint, in 2023, funded the launch of the flagship platform. And in June 2025 it announced a Series B of more than $55 million led by Battery Ventures, with Redpoint, Benchmark, 01A, New View Capital, and World Innovation Lab along for the ride. That brings the total to somewhere around $100 million.
The detail worth lingering on is the lead investor. Battery Ventures' Neeraj Agrawal, who took a board seat as part of the round, is the same partner who years ago led an early investment into Coupa - the incumbent procurement giant that Levelpath is, quite openly, trying to displace. When the person who helped build up the incumbent writes a check to the challenger, that is not proof the challenger wins. But it is a reasonably strong signal about which way the wind is blowing.
What it actually does
Levelpath unifies the pieces of procurement that, at most companies, live in a scattered mess of forms, email threads, spreadsheets, and legacy tools nobody wants to open. Four pillars:
Intake
The front door. Someone at the company needs to buy something; Levelpath gives them a guided, consumer-grade request flow that figures out who needs to approve it and routes it there, instead of the traditional method of asking around until someone in finance sighs.
Sourcing
Running RFPs and collecting supplier bids, with AI that can draft a sourcing event in seconds and analyze the proposals that come back. This is where the "days to seconds" claims live.
Contracts
Reading the unstructured, often deliberately dense text of contracts to surface renewals, risks, and places to save money - the kind of review a human can do but would really prefer a machine handle.
Agents
The 2025 addition: autonomous AI agents that don't just summarize the work but do it, end to end. Levelpath describes procurement as full of "ugh" work, which is an unusually honest bit of product marketing, and the agents are aimed squarely at the ugh.
Does it work?
The most concrete number floating around comes from GATX, the rail-and-asset-leasing company, which reported cutting supplier bid analysis from days to minutes, shrinking RFP production to seconds, and saving $3.5 million from contract analysis. One should always read customer-supplied metrics with a raised eyebrow - they are, after all, selected by the vendor - but the direction is the point. The named customer list is real and not small: Ace Hardware, Amgen, Coupang, SiriusXM, Zendesk, Symetra, C.H. Robinson, Verifone. These are companies with genuine procurement complexity, which is exactly the crowd a legacy-displacement play needs to win.
Who it is up against
The incumbents are Coupa and SAP Ariba - large, established, and, in Levelpath's telling, showing their age. The newer challengers include Zip (valued at $2.2 billion) and Oro Labs. The bet Levelpath is making is not that procurement needs one more feature; it is that the whole experience can be rebuilt around AI and mobile from the ground up, and that being genuinely pleasant to use is itself a moat - because in enterprise software, adoption is the thing everyone struggles with, and adoption is a design problem as much as a technical one.
Whether that thesis holds is one of the more interesting open questions in enterprise software right now. But the ingredients are unusually strong: founders who have won this exact game before, a category everyone agrees is broken, roughly $100 million in the bank, and a lead investor who knows the incumbent's playbook from the inside. Levelpath is, at minimum, a serious attempt to make the software behind the second-biggest line item on the corporate ledger something people don't dread. That is a smaller ambition than "transform the enterprise." It might also be a more useful one.