Databricks Acquires Mooncake (www.databricks.com)

Databricks has acquired Mooncake Labs and is folding its technology into Lakebase — a Postgres-based OLTP database architecture optimized for AI agents. Mooncake enables transactions, analytics, and AI workloads to run on the same fresh data by mirroring Postgres changes to the Databricks Lakehouse in real time, eliminating brittle ETL pipelines and data silos. The integration ties Lakebase directly into Databricks’ Lakehouse and Agent Bricks, giving developers a single foundation for applications, analytics, and agent-driven workflows. This matters because AI agents create and modify data, schemas, and code at machine speed, which traditional OLTP systems and separated analytics platforms aren’t built to handle. Mooncake’s innovations (real-time change capture, Postgres internals expertise, and support for open table formats like Iceberg) promise lower latency, stronger data governance, and fewer operational failures when agents spin up new tables, events, or workflows. With founders experienced in distributed systems and Postgres internals, Databricks is betting this combined stack will reimagine operational databases for the AI era—enabling more reliable, responsive agent-driven applications without the overhead of batch pipelines.