Ottimate.

Who they are now

Picture an accounts payable team at 4:47 p.m. on a Thursday. There is a stack of PDFs in someone's inbox. There is a controller refreshing a screen. There is an invoice from a vendor whose name nobody can spell, for a quantity that doesn't match the purchase order, with a price that changed two cents last week. Somebody has to type all of that into a ledger. Somebody always has.

Ottimate is what happens when that somebody is finally not a human.

The company - a San Francisco fintech with about 270 employees and a wardrobe of finished credentials including Y Combinator, FTV Capital, and a ten-year track record under its old name - sells a single, focused idea: an AI built specifically for accounts payable. It captures the invoice. It reads the line items. It codes them to a general ledger. It routes the document through approvals. It pays the vendor through a network that hands back cash rebates. And it does all of this while a finance team does roughly nothing.

Ottimate is not glamorous fintech. It is not crypto, and it is not consumer. It is the kind of software that nobody outside an office of the CFO is supposed to find interesting. Which, in fairness, is exactly the problem it set out to solve.

The problem they saw

In 2014, three founders - Shishir Deliwala, Bhavuk Kaul, and Ram Jayaraman - went looking at restaurant back offices. What they found was less a business process and more an archeological site. Crumpled invoices in drawers. A bookkeeper with a calculator from 2003. A line item for "tomatoes" that nobody could reconcile to the line item for "produce" entered the week before.

Restaurants are unusually punishing on accounts payable because the line items matter. You can't just file an invoice from a meat distributor under "food." Somebody, somewhere, needs to know how much was spent on ribeye versus chuck so the menu math holds up. Most AP automation tools at the time stopped at the invoice header. Plate IQ - as Ottimate was then called - decided to go further. It would read every line.

That was the bet, and it was, for a while, slightly absurd. Reading line items off a restaurant invoice in 2014 meant fighting OCR engines that could barely tell a "5" from an "S." But the founders kept training the models on real invoices from real kitchens, and the models kept getting less bad, and one day they were not bad at all.

The boring problem turned out to be the moat.

The founders' bet

Y Combinator backed them early. GGV led a smaller round. Then, in November 2021, FTV Capital wrote a check for $160 million in a Series B that, on paper, looked like a vote for restaurants surviving the pandemic. In practice, it was a vote for something else: the idea that the AI Plate IQ had trained on kitchen invoices was general-purpose enough to migrate.

The founders had been quietly working on that migration for years. A construction firm has invoices, too, and the line items - cement, rebar, drywall, hours of labor - are not so different in shape from a wine distributor's. A hospital has invoices. A grocery chain has invoices. A country club has too many invoices. Every CFO who had ever stared at a stack of paper had, in some sense, the same problem.

By December 2023, the company shed the restaurant association entirely. Plate IQ became Ottimate - a name that, with some squinting, rhymes with "automate" and serves as both noun and verb. The tagline is, mercifully, self-aware: Don't just automate AP. Ottimate it.

The product

Strip away the brochure, and Ottimate is four jobs glued into one workflow.

The first job is capture. Invoices arrive however vendors send them - email attachments, paper, EDI files, sometimes a photo from a delivery driver's phone - and the AI extracts every line, not just the total. The second job is coding. Each line gets matched to the right general ledger account using rules the AI learns over time. The third job is approval, which Ottimate routes through whatever workflow the customer has decided is appropriate, with mobile sign-off so a CFO can rubber-stamp from an airport gate. The fourth job is payment, through a network called VendorPay that handles ACH, virtual cards, and checks, and quietly returns cashback on the card volume.

None of these jobs are individually new. AP automation has existed for decades. What Ottimate did - and what its decade of restaurant data gave it - was line-item accuracy at a level the rest of the market struggled to match. When the AI is reading the actual goods, not just the invoice header, the rest of the workflow can be trusted to run on its own. Three-way matching, fraud detection, duplicate flagging, discrepancy alerts - all of it depends on knowing what is actually on the invoice.

The proof

Ottimate runs invoices for thousands of customers across hospitality, healthcare, grocery, construction, retail, and real estate. The hospitality roster - hotels, restaurant groups, country clubs - still anchors the business, but the post-rebrand strategy has been to push into industries where back offices are even less digitized than restaurants were ten years ago. Construction was the first beachhead. Healthcare and multi-location retail followed.

The integrations list is the other tell. Ottimate plugs into QuickBooks, Xero, Sage Intacct, Oracle NetSuite, Acumatica, and a long tail of vertical accounting systems. None of that work is glamorous, and all of it is necessary; an AP automation tool that doesn't talk to your general ledger is essentially a very expensive PDF reader.

G2 has named it a Leader in AP automation across multiple quarters. The Acumatica marketplace lists it as a certified app. Construction trade press declared it the top AP automation product for the sector in early 2024. None of these are Nobel Prizes. All of them, taken together, suggest that the boring parts of the business are working.

The mission

Ottimate's stated mission is to automate the entire invoice lifecycle so finance teams can focus on what actually drives profitability. This is the kind of mission statement that sounds like it was written in a conference room, because it was. But underneath it is something a little more interesting.

Most companies underestimate how much of their finance team's time goes to data entry rather than analysis. The controller you hired to forecast cash flow is, on a Tuesday afternoon, probably re-typing the line items off a vendor invoice. The AP manager you hired to manage vendor relationships is chasing approvals. The CFO you hired to allocate capital is signing PDFs. Ottimate's argument is that the bottleneck in modern finance is not strategy or judgment - it is the mountain of low-value clerical work that has to happen before strategy and judgment can begin.

Remove the mountain, the argument goes, and finance teams become what their job titles always claimed they were.

Why it matters tomorrow

AP automation is one of the corners of fintech where the AI hype actually lines up with the underlying business. Reading documents, extracting structured data, and routing decisions are exactly the tasks large language models and computer vision have gotten good at in the last two years. The companies with the best training data - the most invoices, in the most formats, across the most industries - will compound their accuracy advantage faster than competitors can catch up.

Ottimate has been collecting that data since 2014. The newer entrants pitching AI-native AP have, in many cases, been collecting it since 2023. That is not a fair fight.

The strategic question is no longer whether the AI can read invoices. It is what Ottimate does with the layer above the invoice - the analytics, the spend visibility, the vendor benchmarking, the cash-flow optimization. The product is creeping in that direction. So is the company's tone. The board added Andy Roberts as chairman in 2024, the kind of move you make when you are starting to think about a finish line rather than just a runway.

Back to that AP team at 4:47 on a Thursday. The PDF stack is gone, or rather, it never landed in a human inbox at all. The controller is reviewing exceptions instead of typing line items. The CFO is looking at a dashboard of where the company's money actually went last week, not waiting two months for someone to total it up. Somewhere, a server reads an invoice for tomatoes and codes it to the right account, and nobody notices, which is the point.

The room is quieter. The work is finished. Ottimate did it.