RAMP IS NOW WORTH $44 BILLION. IT IS A CORPORATE EXPENSE CARD COMPANY.
Let me describe Ramp to you. It is software that helps companies manage corporate expenses. Your employees get a corporate credit card. The software tracks what they spend. It helps the finance team see where the money is going, catch waste, and automate the boring parts of running a finance department. That is, in broad strokes, what Ramp does. It is genuinely useful software. Real businesses use it. Real money gets saved. But it is still, at its core, expense management software.
Ramp just raised $750 million at a $44 billion valuation.
For context, that puts Ramp roughly on par with General Mills, the company that makes Cheerios, Pillsbury, and Haagen-Dazs. It is now worth more than Delta Airlines, which flew around 200 million passengers last year. Marriott International, which owns or manages more than 9,000 hotels in 140 countries, is worth somewhere in the $70 billion range. And Ramp, the corporate card company, is priced at $44 billion. Ramp has no hotels. Ramp has no planes. Ramp has a very good expense management platform and an AI story, and in 2026, that is apparently enough.
The round was led by ICONIQ Growth, GIC, and the Ontario Teachers’ Pension Plan, which means it is literally Canadian teachers’ retirement money going into corporate expense software. Goldman Sachs Alternatives, Morgan Stanley Investment Management, D.E. Shaw, Insight Partners, and Generation Investment Management also joined. These are not reckless early stage speculators. These are extremely serious institutions that looked at this company at length and decided $44 billion was a fair entry price.
To be fair to Ramp, the underlying business is real. They crossed $1 billion in annualized revenue. They are free cash flow positive. Their purchase volume grew 170% year over year in March 2026, their fastest growth rate in three years, while operating at roughly 20 times their previous scale. The median customer is reportedly saving 50% more dollars and 32% more hours annually compared to a year ago. Those are not manufactured metrics.
The AI angle is also real. Ramp has been building AI into expense categorization, budget forecasting, vendor negotiation, and financial reporting. The pitch is not just a buzzword slapped on legacy software. They appear to be genuinely rebuilding how corporate finance operations work from the ground up.
But here is the honest version of events. Ramp’s revenue roughly tripled over the past couple of years. Its valuation went from around $16 billion in 2024 to $44 billion today. Revenue growth did not produce that valuation gap. The AI story did. And that gap between growth rate and valuation expansion is exactly the gap that creates uncomfortable conversations when a company finally has to go public and face investors who want traditional financial metrics. From $44 billion, the next stop is either a very large IPO or a very awkward down round. Ramp is betting on the former.
Source: TechCrunch
JEFF BEZOS IS PUTTING $100 MILLION INTO A STARTUP TRYING TO COPY THE HUMAN BRAIN AND USE IT TO FIX THE AI POWER CRISIS
This one is genuinely interesting. Not interesting in the way that every AI press release describes itself as groundbreaking. Actually interesting, in the way that makes you stop and think about what is really being attempted here.
A company called Flourish just closed $500 million in funding to build what it calls Cortex AI, an artificial intelligence system modeled not on how we think the brain works, but on how the brain actually works at the cellular level. The field is called connectomics. Scientists use electron microscopy to map real brain tissue at the level of individual neurons and their connections. You end up with a three-dimensional map of which cells connect to which other cells, how strong each connection is, and how signals flow through the entire network. Flourish’s thesis is that if you understand that architecture deeply enough, you can extract the core principles that make biological intelligence so extraordinarily efficient, and then replicate those principles in silicon.
The reason this matters for money people is the power problem. Training a frontier AI model can require more electricity than a small city uses in a month. Running inference at scale requires server farms drawing power around the clock. Google and Microsoft are reportedly contracting with nuclear power plant operators specifically to feed their AI operations. This is not a minor operational challenge. It is a hard ceiling on how fast AI can grow and how cheaply it can be deployed.
Flourish’s Cortex AI is designed to operate on somewhere between 20 and 50 watts. The power draw of a laptop, not a server rack. If that works at meaningful scale, it would be one of the most significant efficiency improvements in the history of computing. Not incremental improvement. Not a better chip architecture. A fundamental rethinking of how intelligence should be implemented in hardware.
The founders have credibility. Thomas Reardon created Internet Explorer at Microsoft, and then went on to found CTRL-labs, a brain-computer interface company that Meta acquired in 2019 for roughly $1 billion. Rob Williams spent years as an S-team executive at Amazon, meaning he was in the small group that reported directly to Bezos and saw how Amazon thinks about long-term technology bets from the inside.
Bezos personally committed close to $100 million. Lux Capital is in. GV, which is Google’s venture arm, is in. Catalio Capital also participated. The $500 million round values Flourish at $2.5 billion. This is early-stage money for a concept that might take a decade to bear fruit. But the combination of the brain-efficiency thesis, the power crisis context, and the caliber of the investors writing the checks makes this one worth watching closely.
Source: TechTimes
AI MUSIC STARTUP RAISES $400 MILLION AND DOUBLES ITS VALUATION TO $5.4 BILLION WHILE BEING SUED BY THREE MAJOR RECORD LABELS OVER 61,000 SONGS
There is a specific kind of confidence involved in raising $400 million while actively being sued by the three biggest music companies on earth. Most startups would fold. Most founders would settle for whatever terms the labels offered, take their lumps, and quietly move on. Suno raised $400 million.
Suno is the AI music company that lets you type a text prompt and get back a full song. You describe something, the AI writes the lyrics, picks the genre, generates the vocals and the instrumentation, and delivers you a complete track in roughly 30 seconds. It works. More than 2 million people are paying for subscriptions. The company was running at around $300 million in annualized recurring revenue heading into this round. That is a real business with a real product and real customers.
The legal situation is also real. Warner Music Group, Universal Music Group, and Sony Music Entertainment filed suit in 2024 claiming Suno trained its AI on copyrighted music without permission or compensation. The complaint covers 61,000 allegedly infringed songs. These are not bluff lawsuits. The three major labels together control somewhere around 70% of the world’s recorded music catalog, and they have the resources to fight for years.
Here is where it gets interesting. Warner Music Group settled with Suno in November of last year and announced a licensing partnership. One of the three original plaintiffs did the math and decided it was better to get a revenue share from the upside than to keep paying lawyers to litigate a case whose outcome is genuinely uncertain. WMG is now a partner. UMG and Sony are still in the fight, but the trajectory is clear enough that Bond Capital led a $400 million Series D, with IVP, Forerunner, Union Square Ventures, Alkeon, and Quiet Capital joining the round.
The valuation went from $2.45 billion in November 2025 to $5.4 billion today. Six months, one settlement, one label partnership, and the company is worth more than twice what it was. The market has decided the legal risk is manageable and the upside is worth betting on.
The bigger story here is about AI and copyright across the entire industry. Suno now has enough capital to litigate aggressively or settle on favorable terms, depending on what the legal landscape looks like in six months. The outcome of the remaining cases sets a precedent that matters for every other AI company that trained on creative work. $400 million buys you real optionality in a fight this consequential.
Source: TechTimes
DATABRICKS IS QUIETLY BECOMING ONE OF THE MOST VALUABLE COMPANIES IN AMERICA AND IS NOW EYEING A $175 BILLION VALUATION BEFORE IT EVER GOES PUBLIC
Quick question. What is Databricks? If you do not work in enterprise software or AI infrastructure, the honest answer is probably nothing. That is perfectly understandable. Databricks does not have a consumer app. It does not advertise on television. What it does is build the data infrastructure that large companies use to store, process, and run AI on their proprietary information. When a major bank wants to train an AI model on its transaction history, or when a pharmaceutical company wants to run AI analysis on clinical trial data, or when a retailer wants to forecast inventory with machine learning, they almost always need a platform that handles the data preparation, the compute management, and the model deployment at scale. That is Databricks.
Now it is in talks to raise a new round at a valuation of $165 billion to $175 billion. For context, they were at $134 billion in December 2025, just six months ago. Snowflake, the closest public company comparison, currently trades at around $60 billion. Databricks is targeting nearly three times that valuation while remaining private.
The numbers behind the valuation are not fictional. As of February 2026, Databricks was running at $5.4 billion in annualized revenue, growing at more than 65% year over year. They have more than 650 enterprise customers paying over a million dollars per year. Their AI-specific product revenue exceeded $1.4 billion, representing about 26% of total revenue, and growing faster than the rest of the business. And they went free cash flow positive in 2025, which in the current AI funding environment is genuinely remarkable. Everyone else is burning cash and hoping they can raise the next round before the runway runs out. Databricks is running a real business.
CEO Ali Ghodsi has told investors privately that an IPO is coming, most likely in 2026 or 2027. The company added $1.8 billion in debt financing from JPMorgan in January 2026. The institutional investor base includes NVIDIA, Andreessen Horowitz, Morgan Stanley Investment Management, and Teachers Venture Growth. This is not a group doing speculative early stage bets. These are serious capital allocators making a deliberate judgment about enterprise software fundamentals.
When Databricks files its S-1, it will be one of the most carefully read documents in enterprise software history. It will show the public markets what AI-driven data infrastructure actually looks like as a durable business, and it will set the pricing benchmark for every other enterprise AI company considering going public. The people building AI applications on top of enterprise data need Databricks. The bigger the enterprise AI bet gets, the more indispensable Databricks becomes. That is a simple thesis, and the revenue numbers say it is working.
Source: Tech Funding News
JPMORGAN JUST BACKED AN AI COMPANY THAT LETS WALL STREET SPY ON WALL STREET: ALPHASENSE CLOSES $350 MILLION AT $7.5 BILLION VALUATION
This one requires a tiny bit of context to appreciate, but the punchline is worth it.
AlphaSense is an AI platform built specifically for financial analysts, investment firms, and corporate strategy teams. Earnings call transcripts, SEC filings, broker research notes, regulatory documents, industry reports, expert network call recordings. AlphaSense indexes all of it, lets you search it with natural language questions, and synthesizes answers from across the entire corpus in seconds. If you are a portfolio manager who needs to know what every major semiconductor company said about supply chain constraints across their last eight earnings calls, AlphaSense gets you that in under a minute instead of the twelve hours a junior analyst would spend doing it by hand.
The company just raised $350 million in a round led by Vitruvian Partners, with J.P. Morgan Asset Management, Accenture Ventures, D.E. Shaw, and Pinegrove Opportunity Partners participating. The round values AlphaSense at $7.5 billion.
Now here is the thing. J.P. Morgan Asset Management is one of the largest investment managers on earth, running multiple trillions of dollars. They just invested in a company whose software is used by financial analysts to do their research faster and better. Which means JPMorgan put money into a product that helps people who compete with JPMorgan do their jobs more efficiently. That is either a sign of extraordinary confidence in the product’s quality, or evidence that the AI market has reached a point where even the institutions being disrupted are writing the disruptors checks. Probably some of both.
D.E. Shaw is also worth noting here. They are one of the most mathematically sophisticated quantitative hedge funds on earth. They do not make investment decisions based on soft signals. When D.E. Shaw puts money into an AI research tool, it is a reasonable indicator that the product actually delivers an edge worth paying for.
AlphaSense reportedly was already on a path toward profitability before this round. The $350 million is not a survival raise. It is acceleration capital for a company that is already winning. They are betting that the next wave of AI adoption in financial services favors deeply integrated, domain-specific tools over general-purpose chat interfaces. That is a thesis worth paying attention to.
Less flashy than the foundation model labs. No splashy consumer product. But this is the kind of company that files a quiet S-1, prices conservatively, and turns out to be worth dramatically more than the IPO valuation within three years. Watch this one.
Source: Crunchbase News