Ramp announced on October 15 that it had raised $1.85 billion, four months after a smaller round that valued the business at a fraction of what it is worth today. Investors competed for space in the deal, according to people familiar with the matter, and the round closed at a price well above the one discussed when fundraising began.
The new financing values the expense management company at $60 billion, up from $44 billion in the previous round, when it raised $750 million. The round was led by Dragoneer, Thrive Capital, Founders Fund, Sands Capital and Avenir Growth. Ramp has now raised more money in four months than most enterprise software companies raise in a decade, and it has done so while the broader market for late-stage software deals has been uneven.
The company said its annualized revenue has passed $2 billion, up from $1.5 billion in June. It said annualized purchase volume on its platform exceeds $275 billion, an increase of 175 percent from a year earlier. Ramp counts 85,000 customers, among them Visa, Uber, Shopify, Cursor and Figma, a list that mixes companies large enough to have long-standing procurement departments with newer firms that grew up without them.
The product has moved well beyond corporate cards. Ramp has embedded AI agents into purchasing, expense reporting, accounting and budgeting, letting software reconcile receipts, flag duplicate vendors and draft budgets without a person opening a spreadsheet. It has also issued a corporate card designed for AI agents to spend against, an arrangement that assumes software, not employees, will be the party making a growing share of routine purchases.
Analysts said the growth in purchase volume is the number to watch, because it is a measure of how much money actually flows across the platform rather than of how many accounts have been opened. A company that processes $275 billion of annual purchasing earns money on the interchange and on subscriptions, and it also learns where a customer’s spending is going in a way no accounting system does. That data is what makes the agent products plausible and what makes the valuation defensible if growth continues.
The comparison with the previous round is what makes the new one notable. Four months is a short interval for a private company to add $16 billion of valuation, and the willingness of five investors to lead a round of this size suggests they expect the business to reach a scale where the price looks reasonable in hindsight. It also suggests they are competing for allocation, a dynamic that has historically preceded steep corrections when it ends.
The risks are visible in the numbers. Purchase volume growth of 175 percent cannot continue at that rate for long before the base makes it arithmetically difficult. Interchange revenue is exposed to regulation in several markets and to the willingness of merchants to keep paying fees. And the agent-based card product assumes that corporate controls, audits and fraud systems can be adapted to a world in which the spender is not a person, a change that finance teams have not yet been asked to make at scale.
The competitive field is crowded. Brex, Bill and a long list of banks sell overlapping products to the same finance departments, and each of them is adding automation. Ramp’s advantage has been speed of product release and a willingness to take on tasks that traditional providers treat as out of scope. Whether that advantage survives a downturn, when customers scrutinize fees and switch costs differently, is the question the company has not yet had to answer.
The customer list carries an argument of its own. Visa is a payments network that competes with the rails Ramp depends on, and its presence as a customer suggests the company has found a position adjacent to the incumbents rather than directly against them. Shopify and Figma come from the newer generation of software companies that expect finance tools to be programmable, and Cursor, a company that sells AI coding tools, is the kind of buyer whose own spending patterns resemble the automation its vendors are building.
The card issued for AI agents is the more speculative product. Letting software spend against a corporate account requires controls that identify which agent is acting, under what limits, and against which budget, and it requires an audit trail that satisfies accountants who have never reviewed a transaction authorized by a model. Finance teams have not been asked to accept that arrangement at scale, and the early users of such products tend to be companies with few legacy controls.
Ramp has not said how the new capital will be divided between hiring, product development and the infrastructure its agents run on. Nor has it said whether the automation products are sold as part of the standard contract or as a separate subscription. Those choices will determine whether the company’s revenue resembles that of a payments processor or that of a software vendor, which is the difference between a business valued at a multiple of volume and one valued at a multiple of profit.
Ramp has not disclosed its losses, its gross margin, or how much of the new capital will be spent on hiring rather than on buying capacity for the agents it now runs. The company has also not said what portion of its 85,000 customers use the automation products rather than only the card. Those details would say a great deal about whether the valuation reflects a payment business, a software business, or both at once, and the company has chosen to let the top-line numbers make the argument instead.
