Two Companies. Different Sectors. The Exact Same Playbook.
Ramp announced it had raised $750 million in a Series F round at a $44 billion valuation — led by ICONIQ, GIC, and the Ontario Teachers' Pension Plan, with Goldman Sachs Alternatives, D.E. Shaw, and Morgan Stanley Investment Management joining as new investors. Total equity raised by the company since its 2019 founding now exceeds $3 billion.
A year earlier, in June 2025, a healthcare technology startup called Tennr raised $101 million in a Series C led by IVP, with participation from Andreessen Horowitz, Lightspeed, Google Ventures, and ICONIQ. Tennr's valuation hit $605 million. Its product: AI that reads and processes the faxes that healthcare providers still use to refer patients to specialists.
Ramp versus Tennr. Corporate finance versus healthcare administration. A $44 billion giant versus a fast-growing upstart. It would be easy — and completely wrong — to treat these as unrelated stories from unrelated sectors about unrelated investor theses.
Strip away the sector labels and what you find is a single blueprint, executed twice, in two different industries, yielding two very different cheque sizes but the exact same investment logic. Understanding that logic is the most practically useful thing a B2B founder can do right now — because it is not just the blueprint for what got funded. It is the blueprint for what is getting funded next.The Problem With "AI Company" as a Category
Before the blueprint, a clearing: neither Ramp nor Tennr succeeded by positioning themselves as AI companies.
Ramp is a corporate spend management platform. It started as a smarter corporate card and has grown into what its CEO calls "the financial operating system" for modern businesses — one that now manages expense reports, procurement workflows, accounting integrations, vendor payments, and, most recently, AI token spend for companies whose biggest new cost is the intelligence they are buying from OpenAI and Anthropic.
Tennr, for its part, has been deliberately careful about how it describes itself. "I want to talk about problems and I want to talk about solutions," CEO Trey Holterman told Fortune at the time of the raise. "I don't want to talk about just the technology."
That framing is not accidental. It is the first tell in the blueprint. Both companies solved for a specific, painful, expensive problem that existed long before AI arrived — and used AI as the mechanism to solve it, not as the product itself.
This distinction matters more than most founders realise. The venture capital graveyard of 2023 and 2024 is full of companies that built AI wrappers around existing workflows, dressed them up as transformative, and discovered that investors who had been through the SaaS boom were not prepared to fund "AI-powered X" in the abstract. What they would fund — and did fund, at increasing scale — were companies that could point to a specific workflow, a specific buyer, and a specific dollar figure that disappeared when the product worked.
What Ramp Actually Solved — and Why It Compounded
Ramp's founding insight was deceptively simple: corporate spending is broken, and the tools built to manage it — legacy expense platforms, clunky corporate cards, disconnected accounting software — were designed to record spend after it happened rather than control and optimise it in real time.
The company built a corporate card that was smarter than a corporate card. It categorised transactions automatically, flagged anomalies, integrated with accounting software, and gave finance teams visibility they had never had before. The product saved time and money, immediately, for every company that used it.
That immediate, measurable value is what created the flywheel. The median Ramp customer now saves 50 per cent more money and 32 per cent more time annually compared to a year ago, according to the company's own data from May 2026. Customers using the full product suite report even greater gains. Total payment volume grew 170 per cent year-over-year in March 2026 — the company's highest growth rate in three years, despite being roughly 20 times the size it was when it first started growing that fast.
"We're growing as fast as we were three years ago, at roughly twenty times the size," said Eric Glyman, co-founder and CEO of Ramp, in the Series F announcement. "And that's because finance is going through the biggest structural change since the spreadsheet."
The valuation trajectory tells the underlying story in numbers. Ramp was valued at $16 billion in June 2025. By November 2025 it was $32 billion. By June 2026 it was $44 billion. A 175 per cent increase in twelve months. This is not a company that got lucky on a hot AI narrative. It is a company with over $1 billion in annualised revenue, positive free cash flow, 70,000 customers including Visa, Uber, Shopify, Anduril, Figma, and Notion, and 3,200-plus enterprise customers generating over $100,000 in annualised revenue each.
The capital followed the proof. Every time.
What Tennr Actually Solved — and Why the Fax Machine Is the Point
Tennr's founding insight is arguably more counterintuitive than Ramp's, which is exactly why it is more instructive.
Healthcare in America runs, to a degree that would be shocking to anyone outside the sector, on fax machines. When a primary care doctor refers a patient to a specialist, the referral typically arrives by fax, email, or an outdated portal. It must then be manually reviewed, matched against payer-specific documentation requirements, approved or denied, and tracked — all before the patient ever receives care. Each year, more than one in three Americans is referred for specialty care, imaging, or treatment. The process that follows is, according to the company, a "black hole" — patients disappear into a bureaucratic maze, and neither the referring doctor, the receiving provider, nor the patient has meaningful visibility into what is happening.
Tennr built an AI model called RaeLM, trained on more than 100 million deidentified healthcare documents and over 8,000 payer-specific documentation requirements, that reads referrals, extracts relevant information from scanned documents and handwritten forms, routes it appropriately, and automates the workflows that follow. The company now processes 10 million documents a month. It more than tripled its revenue in the two quarters between its Series B and its Series C — which was the reason the Series C closed when it did.
"We're somewhat maniacal as is, but this round is about getting more aggressive in addressing the 'black hole problem' of the U.S. healthcare system," CEO Trey Holterman said at the time of the raise. "It means giving more providers the tools to create a world-class experience for their referral sources and their patients."
The investment thesis from IVP, Andreessen Horowitz, and Lightspeed was not "this is an interesting AI application." It was: "there are hundreds of millions of faxes moving through the American healthcare system every year, this company has built the only model that can read them at scale, and the buyers are healthcare providers who are currently paying humans to do what this product does automatically."
The Blueprint: What These Two Raises Actually Have in Common
Laid side by side, Ramp and Tennr look like companies from different universes. One manages billions in corporate spend. The other reads medical faxes. But the investor logic underwriting both raises is identical, and it can be reduced to five components.
The first is a large, documented, existing cost. Ramp targets corporate spend — a number that every company already has in its budget and every CFO already owns. Tennr targets healthcare administrative burden — a cost that the American healthcare system is actively and loudly trying to reduce. Neither company created the problem. Both found a problem that buyers were already spending money on, and built something that reduced that cost measurably.
The second is a specific, identifiable buyer. Ramp sells to CFOs and finance teams. Tennr sells to the operations staff at specialty care providers, imaging centres, and healthcare organisations. Neither company is selling to "the market" — they are selling to a specific person inside a specific organisation who has a specific line item in their budget that the product addresses directly.
The third is a workflow that was already being done manually. Ramp's early customers were using a combination of corporate cards, expense reports, spreadsheets, and manual reconciliation to manage spend. Tennr's customers were using humans to read and process faxed referrals. Both products replaced a known, staffed, manual process — which meant the ROI calculation for the buyer was not "this might save us money" but "this will reduce X headcount or prevent Y hours of manual work."
The fourth is retention built into the product's structure. Ramp lives inside a company's financial operations. Once it is integrated with accounting software, connected to corporate cards, and embedded in expense approval workflows, removing it is painful. Tennr lives inside the referral operations of healthcare providers. Once it is processing millions of documents a month and connected to provider networks, the switching cost is significant. Both products create stickiness not through loyalty programmes or lock-in clauses, but through operational dependency — they become the infrastructure of the workflow they serve.
The fifth, and most important, is the metrics that confirm the value before the ask. Tennr had more than tripled revenue in two quarters before it raised its Series C. Ramp had crossed $1 billion in annualised revenue before its $750 million Series F. Neither company walked into their raise and told investors what the product might do. They showed what it had already done.
Why This Pattern Is Accelerating — and What It Means for Every B2B Founder
The reason this blueprint is generating increasingly large cheques in 2026 is specific. The AI moment has created a set of conditions that make workflow-level automation suddenly viable at a scale and cost that did not previously exist — meaning that problems which were always expensive enough to solve but previously required too much engineering effort to address are now solvable by teams that are faster and cheaper than the prior generation.
This has caused investors to look at every sector and ask a version of the same question: where is the expensive, manual, regulated, friction-filled workflow that a large number of organisations are currently staffing with humans, that AI can now do better and cheaper? Healthcare administration is one answer. Corporate finance is another. The list goes on: legal document review, insurance claims processing, logistics documentation, revenue cycle management, procurement operations.
The companies that are raising in these spaces are not pitching a technology. They are pitching a replacement for a headcount line that already exists in a buyer's budget. That is a fundamentally different kind of sale — and a fundamentally different kind of funding conversation.
For founders building in any B2B sector right now, the question worth asking is not "what AI features should I add?" It is: "What is the most expensive manual process in this industry, who owns the budget for it, what does it cost them per year, and can I show that my product cuts that cost measurably?" If you can answer all four parts of that question with real data from real customers, you are building the kind of company that the 2026 funding market is looking for. Not because investors have suddenly become generous. But because a workflow problem with a documented dollar cost and a measurable solution is, in any market cycle, the thing that is always worth funding.
Ramp did not raise $750 million because investors believed in spend management as a category. Tennr did not raise $101 million because healthcare AI was a hot narrative. They raised because they had already solved the problem for enough customers that the remaining question was not "does this work?" but "how fast can we scale it?"
That is the blueprint. The capital is the consequence. And the companies that understand that distinction — the ones that build the proof before they pitch the raise — are the ones who will be announcing their own rounds while everyone else is still debating their positioning.
