Databricks just launched Lakewatch, a new AI-powered security product built on the back of two startup acquisitions, according to TechCrunch AI. The company quietly acquired Antimatter in a deal that closed last year and scooped up SiftD.ai in a deal that wrapped up just this Monday.
This is Databricks flexing its $5 billion war chest from last month’s fundraise. The move signals something important: the company isn’t just a data analytics platform anymore. It wants to own the AI security stack.
What Lakewatch Actually Does
Lakewatch combines Databricks’ massive data storage capabilities with classic Security Information and Event Management (SIEM) tasks: threat detection and investigation. The twist: it runs on AI agents powered by Anthropic’s Claude.
SIEM is a crowded market dominated by players like Splunk, Microsoft Sentinel, and CrowdStrike. Databricks is betting that its data lakehouse architecture gives it a structural advantage. It already sits on the customer’s data, so why not secure it too?
The Two Acquisitions
Here’s what Databricks picked up:
- Antimatter — Founded by security researcher Andrew Krioukov, this startup raised $12 million led by New Enterprise Associates in 2022. It built a “data control plane” that let enterprises deploy AI agents securely while protecting sensitive data. Krioukov demoed the tech at RSA’s Innovation Sandbox Contest in 2024. He’s now leading the Lakewatch team at Databricks.
- SiftD.ai — A tiny, very young startup co-founded by Steve Zhang, former chief scientist at Splunk who created the Search Processing Language. SiftD launched its product just last November: an interactive notebook designed for human-agent collaboration. This looks like a classic acqui-hire.
Terms weren’t disclosed for either deal. Antimatter had fewer than 50 employees and SiftD had only a handful, as detailed in TechCrunch AI.
Why This Matters
Three things stand out here.
First, Databricks is making a direct play against Splunk (now owned by Cisco for $28 billion). Hiring Splunk’s former chief scientist and building a competing SIEM product is about as direct a challenge as you can get.
Second, the AI agent angle is real. Security teams are drowning in alerts. An AI-powered SIEM that can actually investigate threats, not just flag them, would be a genuine improvement over current tools. The choice of Anthropic’s Claude as the underlying model is notable, given the enterprise trust Claude has built around safety and reliability.
Third, Databricks signaled it’s not done shopping. A spokesperson told TechCrunch AI: “We’re always looking to what’s next, our goal is to stay ahead of the market and close gaps in what our customers need.”
With billions in revenue and a fresh $5 billion in the bank, expect more acquisitions. Databricks is building an end-to-end platform where data storage, analytics, AI, and now security all live under one roof. For security vendors competing in the SIEM space, that’s a serious new entrant to watch.
Full details are available in the original report on TechCrunch AI.