Databricks Launches Lakebase Postgres on Google Cloud
The fully managed, serverless Postgres database is now generally available on Google Cloud, bringing operational data onto the Databricks platform.
Updated on Nov. 5, 2026 in Data Centers
Databricks has made Lakebase Postgres generally available on Google Cloud. The serverless database runs on the Databricks Data + AI Platform and shares a storage layer with analytical data.
Why it matters
Operational data previously required building and maintaining ETL pipelines, duplicating storage and accepting data lag. Databricks says the consolidation matters more as AI agents act on data in real time.
Vendor-reported figures put multi-zone endpoint uptime at 99.99% versus 99.9% for single-zone endpoints, and Lakebase Search at 2x throughput and 4x lower cost than pgvector. The benchmark methodology and independent verification were not detailed, and private connectivity through Google Cloud Private Service Connect remains in beta.
The players
Databricks
Databricks is a San Francisco-headquartered data and AI company whose platform serves about 20,000 organizations worldwide, including roughly 70% of the Fortune 500.
The details
Lakebase separates compute from storage and uses a Lake Transactional/Analytical Processing architecture, so operational workloads run on the same platform, security and governance as analytics and AI. The offering includes serverless autoscaling, scale to zero, instant zero-copy branching of production data and point-in-time recovery, with access governed centrally in Unity Catalog.
Timeline
November 5, 2026: Databricks announced general availability of Lakebase Postgres on Google Cloud.
Roadmap
Lakebase extends the Databricks lakehouse model by bringing transactional Postgres workloads onto the same storage layer used for analytics. The move positions Databricks against database and data platform rivals as operational and analytical data converge.
Teams running operational databases could consolidate transactional and analytical data onto one platform, reducing ETL pipelines and duplicated storage. Developers gain instant branching of production data and point-in-time recovery, though private connectivity via Google Cloud Private Service Connect is still in beta.
The takeaway
Organizations evaluating where operational data lives should weigh the cost of maintaining separate pipelines against consolidating on a shared storage layer. Checking which connectivity features are production-ready, such as the beta private connectivity option, will shape deployment timing.
Further reading
Explore more Data Centers coverage for context on how cloud database infrastructure is evolving.
Source note: This article includes information reported by HPCwire.
