Databricks has launched Lakebase, a serverless database solution based on PostgreSQL that allows independent scaling of compute and storage resources. It integrates with the Databricks platform, combining transactional and analytical functionalities. Lakebase aims to simplify real-time application and AI workload development by consolidating database management, analytics, and governance. Key features include instant data branching, point-in-time recovery, and unified access controls. It is designed to address the limitations of traditional operational databases, which struggle with modern AI demands. Lakebase offers a managed PostgreSQL service, supporting up to 8TB per instance and featuring pgvector for AI-driven search. It has been in development since June 2025, utilizing technology from the PostgreSQL company Neon and enhanced by the acquisition of Mooncake. Lakebase is available in Autoscaling and Provisioned versions, with the Autoscaling version billed based on usage. It is currently available on AWS, in public preview on Azure, and expected on Google Cloud later this year. SOC2 and HIPAA certifications are projected for early 2026.