On September 9, 2025, Ramp told the market it had crossed $1 billion in annualized revenue at the end of August [Ramp PRNewswire, Sept 2025]. Six years from incorporation. Three founders. One product surface that started as a corporate card and now reaches into accounts payable, travel booking, procurement, and the general ledger.
The number is the headline. The mechanism is more interesting. Ramp is shipping AI agents into the parts of finance work that used to require a junior accountant and a spreadsheet: a Policy Agent that reviews expenses in real time and flags what is out of policy, and Agents for AP that code invoices, route approvals, and process payments [Ramp PRNewswire, 2025]. Early customers are catching three times more out-of-policy spend, the company says [Ramp]. That is the pitch to the CFO buyer in 2025: not a card, not a dashboard, but software that does the work.
The bet
Ramp sells one platform that bundles corporate cards, expense management, accounts payable, travel, procurement, and accounting automation [Ramp]. The wedge is still the card, which generates interchange revenue on every swipe and gives Ramp a reason to live inside a customer's spend data from day one. The expansion is everything else. Bill pay puts Ramp in the middle of vendor payments. Travel adds a booking surface where an AI price monitor rebooks hotels when rates drop by $50 or more [Ramp]. Procurement moves Ramp upstream of the purchase, before the card ever gets pulled.
The customer base now spans more than 45,000 businesses according to Fortune [Fortune, 2025], with Lightspeed citing over 50,000 companies in its own write-up of the latest round [Lightspeed, 2025]. The mix skews toward startups and mid-market finance teams, the segment most willing to rip out a legacy expense tool for something that promises to eliminate work rather than digitize it.
Why it could be big
The capital tells you how investors are pricing the thesis. Ramp raised $150 million in April 2024 [Reuters, April 2024], then $300 million led by Lightspeed at a $32 billion valuation in 2025 [Crunchbase, 2025], then another $500 million led by Iconiq weeks later, this time pegged at $22.5 billion [Crunchbase, 2025]. The valuation step-down between the two 2025 rounds is unusual and worth flagging, but the fact pattern is clear: two of the most disciplined growth investors in fintech wrote nine-figure checks within the same year.
Series D+ Apr 2024 | 150 | $M
Series D+ 2025 | 300 | $M
Series E-2 2025 | 500 | $M
The cap table reads like a who's who of growth-stage fintech conviction: Lightspeed, Iconiq, Founders Fund, Thrive Capital, and Khosla Ventures [Crunchbase]. Thrive's involvement is notable given Josh Kushner's broader fintech portfolio [The Information]. The thesis these investors are underwriting is that spend management is the wedge into a much larger finance-operations budget, and that AI agents collapse the cost of serving each customer enough to defend gross margins as Ramp moves into bill pay and procurement, where competitors like Bill.com and Coupa already sit.
The team and traction
Eric Glyman (CEO) and Karim Atiyeh (CTO) [LinkedIn, 2026] met at Harvard and previously built Paribus, a price-tracking app that Capital One acquired in October 2016 [TechCrunch, October 2016]. Both joined Capital One as Senior Directors in U.S. Card after the deal, which is the rare founder background that includes operating experience inside a top-five U.S. card issuer. Gene Lee, the third co-founder, focused on growth engineering [Crunchbase/Forbes]. The Paribus DNA is visible in the product: Ramp's hotel rebooking feature is essentially the Paribus refund mechanic applied to corporate travel.
Glyman has been public about the company's AI conviction. In a January 2025 Bloomberg interview, he framed the agent push as the natural next layer on top of Ramp's transaction data [Bloomberg, January 2025]. The September 2025 ARR milestone followed eight months later.
The honest counterfactual
The bear case is real and named. Brex, the most direct competitor, has spent years building a parallel stack with its own AI features and a deeper push into banking. American Express owns the enterprise card relationship and has the brand and float advantage that no startup can match. If Ramp's expansion into bill pay and procurement stalls, the card business alone has to justify a $22.5 billion valuation, which implies a revenue multiple that leaves little room for a slowdown. The valuation reset between the Lightspeed and Iconiq rounds, from $32 billion to $22.5 billion within weeks [Crunchbase, 2025], suggests the market itself is recalibrating what fintech multiples should look like in an AI-agent world.
The bull answer, supported by what Ramp has disclosed: the company hit $1 billion in annualized revenue in August 2025 [Fortune, September 2025], and the agent products are designed to expand share of wallet inside the existing 45,000-plus customer base [Fortune, 2025] before depending on net-new logos. If even a meaningful fraction of those customers move bill pay and procurement onto Ramp, the unit economics get easier, not harder.
What to watch
Three things over the next twelve months. First, whether Ramp publishes durable metrics on agent adoption: how many of its 45,000-plus customers actually turn on Policy Agent or AP Agent, and what that does to net revenue retention. Second, whether the next funding round, if there is one, holds the $22.5 billion mark or pushes higher, which would signal that Iconiq's entry price was the floor rather than the ceiling. Third, whether Brex responds with its own agent layer, or cedes the AI-finance positioning to Ramp.
The deeper question for the reader: if AI agents really do collapse the cost of finance operations, does the winning company in spend management end up looking more like a software business or more like a bank? Ramp is currently betting on the former. Iconiq just paid $500 million to agree.