This article first appeared on GuruFocus.
Databricks just snapped up Tecton, a startup built by ex-Uber engineers who once designed the ride-hailing giant’s in-house AI system. Tecton specializes in feeding machine learning models live data the stuff that makes AI agents answer quickly instead of lagging.
CEO Ali Ghodsi says the acquisition is all about speed. Humans hate to wait, he told Reuters, noting that real-time response is critical for things like voice assistants and other user-facing AI tools. By plugging Tecton into its Agent Bricks platform, Databricks hopes to make its AI agents faster, smoother, and more useful in the enterprise world.
The two companies already knew each other well they’ve partnered since 2022, and many of Tecton’s customers, like Coinbase (COIN, Financials), already use Databricks. Now those connections get tighter.
The timing also matters. Databricks just inked a term sheet valuing the company at over $100 billion up 60% in eight months and it’s using that momentum to scoop up startups that add to its AI arsenal. In the past two years, it has bought MosaicML for $1.3B, Tabular to bring in Apache Iceberg’s creators, and serverless database firm Neon for $1B.
Tecton itself was no small fish. Backed by Sequoia Capital, Kleiner Perkins, Andreessen Horowitz, and Bain, it last raised money at a $900M valuation. About 90 employees will now join Databricks.
Bottom line: Databricks isn’t just building a platform it’s assembling an AI toolkit piece by piece. With Tecton, it’s betting that real-time speed is the edge that keeps customers hooked.