The gap between what the AI data platform wanted and what the market offered reveals how much capital is chasing a narrow set of late-stage AI companies. Databricks had no plans to raise when a June report from The Information surfaced its fundraising efforts during the company’s own conference — and the resulting investor pile-on forced a decision about how much stock to sell.
How a News Leak Became a Self-Fulfilling Prophecy
Ghodsi said Databricks was heads down running its June conference when The Information published a story about a planned fundraise. Within hours, his phone was ringing with investors asking for allocations. “The interest level was just insane,” he said. “Just from this select group of investors that we looked at, there was $15 billion of interest.”
Turning away that much demand risks alienating long-term backers. Databricks chose to issue more stock instead. In July, the company announced it had closed a new round at a $188 billion valuation without disclosing the amount. On Thursday, it confirmed the $5 billion figure and bumped the valuation to $190 billion.
Coatue led the round. Blackstone, MGX, multiple T. Rowe Price accounts, and Sixth Street Growth — the firm founded by former Goldman Sachs chief investment officer Alan Waxman — joined as new investors. Roughly two dozen venture firms participated in total.
The Numbers Behind the Demand
Databricks reported $7 billion in annualized run-rate revenue, growing at 80% year-over-year, and said it is cash-flow positive. Its core cloud data warehouse product accounts for $1.5 billion of that run rate and is still doubling annually.
The company’s newer AI products are scaling quickly. Lakebase, a database designed for AI agents that launched in June 2025, has already reached a $100 million revenue run rate. Genie, an AI chatbot that performs on-the-spot business analysis, is “insanely popular,” Ghodsi said.
Why Raise More When You Already Have $20 Billion?
Databricks had raised $20 billion over the previous 20 months, making the question of why it needed more capital a fair one. Ghodsi pointed to three drivers: cloud infrastructure, AI research, and acquisitions.
The company holds multi-billion-dollar commitments with all three major hyperscalers — AWS, Google Cloud, and Microsoft Azure — to support its compute-heavy workloads. Its AI research team has 100 people, a costly and competitive headcount to maintain. And Databricks is actively buying companies. This week it acquired Electric, the maker of PGlite, a lightweight Postgres database that lets agents spin up databases on demand. In June it bought AI cybersecurity firm Panther, and in March it acquired two additional startups. Terms for the Electric deal were not disclosed.
What Happens Next
Databricks’ decision to stay private has become a running joke in Silicon Valley, with observers noting the company has raised so many rounds it is running out of alphabet letters for series naming. Ghodsi told CNBC he still intends to take Databricks public eventually — a near-obligation given the size of its investor roster — but he is in no hurry.
The math favors patience. With $15 billion in unsolicited demand and the ability to set his own terms, Ghodsi can fund AI infrastructure and acquisitions without the quarterly scrutiny of public markets. The real question is whether Databricks’ 80% revenue growth holds as the AI infrastructure market matures and competition from Snowflake, hyperscaler-native tools, and emerging startups intensifies. If growth slows, the $190 billion valuation will face its first real test — and the window for a public offering on favorable terms could narrow.
— David Kim, technology desk, AXO News