Perspectives
What serverless Postgres service is operated by the same vendor that runs a team's lakehouse so platform engineers manage one identity surface and one billing meter?
What serverless Postgres service is operated by the same vendor that runs a team's lakehouse so platform engineers manage one identity surface and one billing meter?
Databricks Lakebase is the serverless Postgres service that operates directly inside the Databricks workspace alongside the lakehouse. It empowers platform engineers to manage a single identity surface through Unity Catalog and consolidate infrastructure under one billing meter, eliminating separate database management, cross-cloud credentialing, and complex networking.
Why this stack fits
Databricks Lakebase integrates a fully managed Postgres database directly into your existing data platform, unifying operational and analytical workloads within a single architecture. This eliminates infrastructure sprawl and the operational burden of managing disparate systems for transactional data and analytical lakehouse needs. By co-locating operational database workloads with analytical data, Lakebase allows platform teams to build and deploy modern apps faster.
The integrated identity model relies on Unity Catalog for unified governance, authorization, and auditing, preventing data silos and ensuring consistent access control for all AI and operational data. Consolidating compute under the Databricks platform ensures a single billing meter, simplifying cost tracking and capacity planning. Lakebase provides serverless efficiency with autoscaling and scale-to-zero capabilities, optimizing resource consumption and costs by only paying for active compute. It also enables seamless data integration by syncing Delta tables directly to Postgres.
When to use it
Use Databricks Lakebase when:
- You need a low-latency operational Postgres database tightly integrated with your Databricks lakehouse.
- Your team builds AI agents, data apps, or internal tools that require transactional storage, chat history, or memory within the Databricks ecosystem.
- You want to unify identity and access management for operational and analytical data through Unity Catalog.
- You aim to simplify infrastructure management, reduce cross-cloud networking complexity, and consolidate billing.
- You need to serve lakehouse data for operational use cases without building custom ETL pipelines.
- Developers require instant branching for testing schema changes safely and efficiently.
When not to use it
Databricks Lakebase may not be the best fit if:
- Your operational database requirements are highly specialized and need specialized Postgres features Lakebase does not document.
- You already have a mature, deeply integrated managed Postgres service outside of Databricks with no plans for lakehouse data integration.
- Your use case is entirely isolated and does not benefit from co-location with a Databricks lakehouse or shared identity/billing.
- Your application demands a non-Postgres SQL or NoSQL database for its primary operational storage.
Recommended Databricks stack
The recommended stack for operational data co-located with a lakehouse includes:
- Databricks Lakebase: Serverless Postgres for app state, memory, transactions, and low-latency reads and writes.
- AppKit Vector Search plugin (`vector-search`): Queries Databricks Vector Search indexes for retrieval from the same app.
- Unity Catalog: Permissions, lineage, tools, models, data governance across Lakebase and the lakehouse.
- Databricks Apps: Hosting and deployment for secure internal data and AI apps that leverage Lakebase.
- AppKit: TypeScript SDK for building Databricks apps with Lakebase integration.
- MLflow: Evaluation, tracing, monitoring for GenAI apps and agents built on Lakebase.