Databricks is within the means of closing a contemporary spherical at a $100 billion valuation, sources confirmed to TechCrunch. The spherical was initially reported by the Wall Street Journal.
A supply acquainted with the deal tells TechCrunch completely the brand new spherical is about $1 billion, and was wildly oversubscribed. Databricks, finest recognized for its information analytics merchandise, shunned promoting much more fairness as a result of it didn’t want money for operations after its as soon as record-breaking $10 billion increase at a $62B valuation in January, in line with the supply. (OpenAI has since squashed the report with a $40 billion increase in March.)
The spherical was co-led by each Thrive and one in all Databrick’s early buyers, Perception Companions, TechCrunch has realized. These two companies led the final spherical, as effectively. The corporate has now raised about $20 billion because it was based in 2013.
This was a main spherical, that means it didn’t embrace staff promoting their shares. Nonetheless, sources near the corporate say Databricks has already had two secondary rounds for workers in 2025. These provides allowed staff to promote as much as 40%, 50%, or as much as 60% of their shares, relying on the dimensions of their holdings.
In each instances, the supply stated, the complete funds out there for the secondary spherical weren’t maxed out, that means staff held onto extra shares than they might have offered. Whereas Databricks clearly isn’t in a rush to IPO, staff have had two latest probabilities to money out shares.
This new spherical, nonetheless, was raised to pursue two particular tasks — a database for AI brokers and its AI agent platform — Databricks co-founder and CEO Ali Ghodsi informed TechCrunch in an interview.
The corporate will make investments closely in its database for AI brokers, making it usually out there to all prospects. It launched the product, often known as Lakebase, in June at its annual tech convention. Lakebase, which is predicated on the open supply database Postres, is enterprise-grade and helps company builders’ vibe coding tasks. This makes it a competitor to Supabase.
Techcrunch occasion
San Francisco
|
October 27-29, 2025
“The database market is $105 billion of TAM, of income, sitting there, type of unaffected within the final 40 years,” Ghodsi informed TechCrunch, giving a refined nod to how database large Oracle has had a lock in the marketplace for many years. TAM refers back to the whole addressable market.
“Right here’s the attention-grabbing statistic no person’s taking note of: a yr in the past, we noticed within the information that 30% of the databases weren’t created by people. For the primary time, they have been created by AI brokers. And this yr, the statistic is 80%,” he stated, including that he predicts this stat to extend to 99% of latest databases inside a yr.
“There’s a brand new consumer. The consumer shouldn’t be human. It’s an AI agent, and if we simply double down on making that consumer persona profitable, that’s the wedge to disrupt that TAM,” he stated.
As for a way Lakebase will differentiate from Supabase and others already constructing Postgres-based databases for brokers, Ghodsi stated the hot button is “separated compute and storage.”
By untying the dear compute from the lower-cost storage, Databricks can affordably let customers create many databases. “As a result of these brokers are tremendous quick. They simply spin up plenty of databases, a lot quicker than people can, however you don’t need to go bankrupt since you’re doing that,” he defined.
The second challenge Databricks will likely be investing closely in is AI agent platform Agent Bricks, additionally launched in June. “All people’s tremendous targeted on tremendous intelligence,” Ghodsi stated. “However that’s not what we’d like in organizations.”
Somewhat than synthetic normal math geniuses or cancer-curing scientists, what corporations want are brokers that may reliably deal with, unaided, mundane duties like onboarding staff or answering personalised questions on HR advantages.
“I believe that’s a a lot larger alternative, truly, for the worldwide GDP and for organizations,” he stated. He believes that such focus will give Agent Bricks a aggressive benefit.
He additionally raised the additional money so Databricks can get into the AI poaching wars. “As you already know, it’s fairly costly to rent AI expertise proper now,” he smiled.