He runs a credit card company whose whole pitch is to get you to spend less. Wall Street thought that was a joke. It is now worth $32 billion.
Every corporate credit card in America ran on the same quiet bargain: spend more, earn more. Points, miles, cash back. The more a company burned, the more the issuer smiled. Eric Glyman looked at that arrangement and decided to build the opposite. Ramp, the company he co-founded in 2019, makes money helping businesses cut their bills. Fewer subscriptions. Cheaper flights. Vendors renegotiated. The pitch was heresy in a rewards-obsessed industry, and it worked so completely that private investors valued Ramp at $32 billion by late 2025.
Today Glyman is not selling a card. He is selling absence - the disappearance of the busywork that eats finance teams alive. Ramp's software chases receipts, fills in expense reports, flags waste, and pays bills before anyone thinks to ask. He calls the destination “self-driving money,” and he means it literally: the way a self-driving car takes the wheel, Ramp's AI agents are meant to take the ledger. “The best AI is invisible AI,” he says. The compliment he is chasing is a customer who forgets the software is even there. It is a strange ambition for a founder - to build something so useful that people stop noticing it - but it is the whole thesis, and it explains almost everything about how he runs the place.
He is also, by his own accounting, a man obsessed with a number. Ramp is not two years old or six years old to him. On any given morning it is day 2,367, or day 2,368 - a specific, numbered, non-renewable unit he refuses to waste. “Today is the only day 2,367 we're going to have,” he told Fortune. “We're going to make it count.”
The idea that made him a billionaire began as a small annoyance. Glyman had bought discounted flights for himself and some friends. The next day, the same fares dropped roughly $100. Most people shrug. He read the fine print, discovered the airline owed a refund, and clawed the money back for the whole group.
That instinct - the fine print is hiding your money, go get it - became Paribus, the startup he launched in 2014 with his Harvard friend Karim Atiyeh. Paribus watched your purchases and automatically filed for refunds when prices fell. It went through Y Combinator, and in 2016 Capital One bought it, folding it into what became Capital One Shopping.
Glyman and Atiyeh could have cashed out and coasted. Instead they spent nearly three years inside the bank, and that is where the real education happened. Glyman watched customers exhaust themselves gaming rewards programs while the banks that issued those rewards were quietly incentivized to devalue them. The whole game, he realized, was rigged against the person holding the card.
A refund robot for ordinary shoppers. Buy something, watch the price drop, get your money back automatically. Sold to Capital One in 2016.
Rewards reward the issuer, not you. So Glyman built a company that profits only when its customers spend less - the exact inverse of the industry.
Karim Atiyeh and Gene Lee. The three left comfortable jobs in 2019 to start Ramp “from a sheet of paper in an apartment.”
“We all know money talks. We're teaching it to think.”ERIC GLYMAN · ON RAMP'S AI
Glyman's bet is that the future of finance software is not a smarter chatbot you talk to. It is software you never have to talk to at all.
He borrows the metaphor of the self-driving car on purpose. A great autonomous vehicle does not ask you which lane to take - it just gets you there. Applied to money, that means the expense report that used to take an hour now takes about ten seconds, and eventually takes zero, because an agent already did it. Receipts get chased. Cash flow gets forecast. Policy gets enforced. Waste gets flagged. None of it requires a human to remember to log in.
It is a contrarian read on the AI moment. While much of the industry raced to bolt conversational assistants onto everything, Glyman argued that in finance the highest form of the technology is the kind you cannot see. “Many tasks that need to be done manually right now,” he says, “will happen without you even asking.”
Agents hunt down missing receipts and match them to transactions automatically.
Expense reports auto-complete. What took hours collapses to seconds.
Cash flow is projected in the background so humans can decide, not calculate.
The best AI is invisible AI. Our agents chase receipts, auto-fill expense reports and forecast cash flow so humans can focus on building things that matter.
Today is the only day 2,367 we're going to have. We're going to make it count.
Our whole mission is to help our customers spend less.
I'm the kind of person that does the top 10 interesting things - but at scale that will kill you.
Most founders cling. Glyman does the opposite on purpose. He talks openly about trying to “put yourself out of a job very, very often” - handing off even the work he most enjoys, because the things one person finds interesting do not scale to a company of thousands. His stated rule: double down on your strengths, hire for your weaknesses.
He keeps unusual company for advice, name-checking figures like Fidji Simo and Satya Nadella. And he runs Ramp at a tempo that has become its own legend: the company claims to ship more product features in a year than there are business days on the calendar. Speed is not a slogan there; it is the metric.
Underneath the velocity is a certain humility. “There's a lot of things I don't know,” he has said - an odd admission from someone whose company grew faster than almost any in New York history, and maybe the reason it did.
It sounds like a paradox that should have bankrupted him. It did the opposite.
The traditional corporate card is a spending engine dressed up as a perk. Issuers hand out points, so finance teams chase points, so spending goes up, so the issuer wins. Glyman's inversion was to align Ramp's incentives with the customer's actual interest. Ramp surfaces duplicate subscriptions, unused licenses, cheaper vendors and pricier-than-necessary flights, then automates the approvals, the reconciliation and the accounting that normally swallow a finance department's week. The company says customers cut expenses meaningfully and close their books many times faster after switching.
The scale of that promise is why the numbers moved the way they did. Ramp reports it has saved customers a collective $10 billion and automated 27.5 million work hours. It grew from a run rate in the tens of millions to more than a billion in annualized revenue, all while serving upward of 40,000 businesses. The market it is chasing - corporate card and spend - is measured in the trillions, and Ramp has captured only a sliver of it, which is precisely the argument its backers make when they mark it at $32 billion.
What separates Ramp from a dozen expense-software startups, in Glyman's telling, is tempo and direction. The company ships at a rate that would be reckless anywhere else, and it points every one of those features at the same north star: take work off the customer's plate. Speed plus a single obsession is a hard combination to copy, and it is the reason a founder who once clawed back $100 on a plane ticket now sits on the Forbes billionaires list.
“Many tasks done manually right now will happen without you even asking.”ERIC GLYMAN · ON SELF-DRIVING MONEY