Databricks just raised $7 billion in equity and debt at a valuation of $134 billion. I don’t blame you if your eyes are cloudy. This is the first mega-round for the company to count, and at some point, it won’t be news for another private company to raise another huge amount of money.
But here’s why it’s important. Databricks is starting to behave more like a public company as it inches closer to its IPO. And the company’s financial statements tell us more than just the size of the check.
80% of databases on the Databricks platform are currently built by AI agents rather than humans. Not all of these are technology companies. Databricks has over 20,000 customers.
This is the clearest evidence yet of what we have been talking about since 2026 began. AI agents aren’t just writing code; they’re building real software inside the world’s largest companies. And that has a huge impact on trade.
We spoke with Databricks CEO Ali Ghodsi. He sits at the intersection of all of that: models, data, infrastructure, and the 20,000 companies that are actually trying to leverage all of that. He knows which AI models are winning, which AI models companies are using, and how quickly agents are improving, giving him a front-row seat to answering the question the entire software industry is currently grappling with: What happens when AI can build artificial intelligence?
