Databricks just pulled a billion-dollar power move — and it’s all about keeping pace with AI agents that don’t sleep, don’t eat, and definitely don’t wait for devs to spin up databases manually.
In a deal worth around $1 billion, Databricks is acquiring Neon, a cloud-native startup building a slick, open-source alternative to AWS Aurora Postgres. Why? Because the future of AI development is fast, serverless, and automated — and Neon’s got the kind of infrastructure that speaks that language fluently.
Let’s break it down: Neon’s platform is built for the age of AI-native development. It offers serverless Postgres with modern features like database branching (so you can test changes without wrecking production), point-in-time recovery, and dynamic scaling of compute and storage — all of it designed to run as effortlessly as the AI agents calling the shots.
And here’s the kicker: 80% of the databases on Neon are spun up by AI agents, not humans. That stat alone had Databricks CEO Ali Ghodsi practically jumping into his wallet. In his words, “Agentic apps are reshaping what a database must do.” So why not bring on a database that can keep up?
This isn’t Databricks’ first AI-forward acquisition either. They scooped up MosaicML last year for $1.3B to boost their large language model (LLM) capabilities, and grabbed Tabular for nearly $2B to supercharge data collaboration. Now with Neon in the mix, they’re building an end-to-end AI dev stack that stretches from data pipelines to real-time inference — and now, to serverless relational storage too.
Bottom line: Databricks isn’t just trying to keep up with the AI boom. They’re actively shaping how it unfolds — one billion-dollar acquisition at a time.
