Databricks Introduces Lakebase Serverless Postgres Database

Databricks announced Lakebase, a new category of serverless Postgres database built for modern software teams, featuring instant branching, separate compute and storage, and an architecture designed to scale with app and agent workloads. The company published a guide, Lakebase For Dummies, covering why traditional databases break down under modern workloads, how separating compute from storage changes cost, speed and reliability, and how the architecture fits alongside a lakehouse for unified transactional and analytical work.
Event version 1.
Key insights
Serverless Postgres category
Lakebase is described as a new category of serverless Postgres database built for the way teams write software today. It includes instant branching and separate compute and storage.
Designed for app and agent workloads
The architecture is designed to scale with app and agent workloads. Databricks positions it against traditional databases that break down under modern app and agent workloads.
Compute-storage separation
Separating compute from storage changes cost, speed and reliability, according to the announced guide. Instant branching and cloning are presented as capabilities for development teams.
Lakehouse integration
The Lakebase architecture fits alongside a lakehouse for unified transactional and analytical work. The guide covers this integration.