The mathematician who decided the month-end close was a solvable problem
EQT Ventures flew to Amsterdam and spent two full days in a room with Albert Malikov before they wrote a cheque. What they came away talking about was not the deck or the demo. It was the obsession - "the level of obsession required to solve a problem of this scale." That is an unusual thing for an investor to fixate on. Most of them want the market size. With Malikov, they wanted to be sure he would not stop.
The problem he refuses to stop on is the financial close. Every month, finance teams across the world reconcile data scattered across ERPs, bank feeds and payment systems - a grind that can swallow up to twelve days. The market for fixing it is worth around $19.5 billion. And for two decades, essentially one incumbent, BlackLine, has owned it. Malikov looked at that and saw not a crowded field but an empty one.
His company is called Stacks. The pitch fits on a business card: AI to close your books in one click. In February 2026, Stacks raised a $23 million Series A led by Lightspeed Venture Partners, with EQT Ventures, General Catalyst and S16VC piling back in. It landed less than a year after the company came out of stealth, and on the back of a seed round that had closed only months earlier. For a category that moves at the speed of a quarterly audit, this is sprinting.
We started with the most manual and foundational workflows in finance: accounting and the close. From day one, we focused on solving the core problem - fragmented data.
- Albert Malikov, Founder & CEO, StacksMoney rails, learned the hard way
Before he was a founder, Malikov was the person companies trusted with their payment infrastructure. At Plaid he ran international product, building the open banking platform that became the leading one in Europe and standing up a vertically integrated team of more than a hundred people through the messy years of finding product-market fit and scaling past it. Before Plaid, he led product for Uber Money - global payment strategy, regulation, identity and verification, risk. These are not glamour assignments. They are the parts of fintech where a rounding error becomes a regulatory filing.
Go back further and the pattern holds. He co-founded inShopper, a cash-back service that was acquired by Mail.Ru Group. He cut his teeth at UBS Investment Bank covering Emerging Markets Equity. The throughline is not a single industry so much as a single instinct: find the place where money moves badly, and rebuild the machine underneath it.
Why finance was last
Ask Malikov why AI swept through legal work before it touched finance and he will give you a structural answer, not a hype one. "In legal, the data is very text-based and consolidated in one place. That's why we see AI penetrating much faster there," he told Lightspeed. "In finance, the data structures are a lot more complex." That complexity is exactly the moat he is digging into. Stacks pulls from ERPs, bank accounts and payment systems and stitches the fragments into something an AI agent can actually reason over - close management and reconciliations, journal entry automation, flux and variance analysis, intercompany and AR modules.
The customers are not pilots. Epidemic Sound, Pleo, Cleo and Bloom & Wild are among the thirty-plus enterprises on the platform, and the ones willing to talk numbers report closing 50 to 60 percent faster. Across its book of customers, Stacks reckons it is handing finance teams back more than 100,000 hours a year. That is the part of the story that travels: the product does not promise to make accountants feel modern, it promises to give them their evenings back.
In legal, the data is consolidated in one place. In finance, the data structures are a lot more complex.
- Albert Malikov, on why AI hit finance lateThe immigrant's compound interest
Malikov trained as a mathematician at Lomonosov Moscow State University, then took an MBA at MIT's Sloan School of Management. His early backers describe someone who immigrated from Russia and built a cross-border career against systemic constraints, carrying what one of them called a resilience and drive to prove himself. It is the kind of biography that explains the two-day Amsterdam interrogation: people who have had to earn every door tend not to walk away from a hard problem just because it is hard.
There is something quietly funny about a man with a master's in pure mathematics deciding that his life's work is the spreadsheet - or, more precisely, the abolition of it. Most people who can do that kind of math run from accounting. Malikov ran toward it, on the theory that the most boring corner of the enterprise is also the one where a well-aimed AI agent creates the most undeniable value. Nobody throws a parade for a faster close. They just stop dreading the first week of the month.
What he is building toward
The ambition is bigger than a tool. Malikov talks about rebuilding enterprise finance operations on agentic AI - automating the manual, error-prone work of accounting so finance teams can stop being data janitors and start being analysts. The one-click close is the headline. The real bet is that once the close is automated, real-time financial visibility stops being a slide in a board deck and becomes the default state of the business. Stacks is moving like a company that believes the window to define this category is open right now and will not stay that way.
For now, the scoreboard is simple. Two rounds in under a year. A roster of enterprise logos. A former Plaid and Uber operator who can talk to a CFO and a machine-learning engineer in the same breath. And an investor base that keeps doubling down because, having spent time in the room, they came to the same conclusion he did: the close is broken, and someone this stubborn might actually fix it.