Most developers are familiar with the hidden costs of traditional databases that tie together computation and storage. This architecture often leads teams to create custom infrastructure for managing developer workflow rather than focusing on building. More importantly, it creates a dangerous resource conflict. Because each query competes for the same fixed CPU and memory resources, a single query can impact all live operations.
These barriers slow down teams and make working against live data risky. As applications become more automated and systems act on data in real time, this type of shared, fragile infrastructure becomes an even bigger liability.
To overcome this architectural hurdle, we created Lakebase RangeA new architecture for operational databases that separates computation from storage. Today, we are pleased to announce that Databricks LakebaseThe first implementation of this category is Generally available on AWS.
For decades, the architectural ‘tax’ of keeping operational and analytical data separate has slowed enterprise innovation. By separating the storage layer and integrating directly with the data lake, LakeBase is positioned as a new class of operational database that treats the infrastructure as a flexible, on-demand service. For companies building advanced AI capabilities, this means there is less chance of manual disruption to the database. It becomes a tool that agents can develop and manage more independently to keep pace with the pace of AI development. This reflects a broader shift toward architectures that reduce data movement and duplication and bring together operational, analytical and AI workloads.-Devin Pratt, research director at IDC
Ship faster with serverless managed Postgres
Lakebase General Availability provides a fully managed, serverless Postgres service with the uptime and predictable performance required for production applications. By separating compute from storage, it automates configuration and resource management tasks that typically slow down development. Its new architecture automatically scales to handle heavy queries, keeps apps responsive under load, and supports instant data branching so teams can safely test and develop without risking production. Since its launch in June 2025, the adoption rate of Databricks’ data warehousing product has more than doubled, with thousands of companies running production workloads directly on their operational data.
Key capabilities available today include:
- Serverless Autoscaling And zeroing the scale: : Dynamically adjust resource counts to match traffic spikes and turn off completely when idle to eliminate wasted costs.
- accelerated database branching: : Create isolated, zero-copy clones of production data in seconds for risk-free testing and development.
- Point-in-time recovery (PITR): Protect against accidental deletions or bugs with millisecond-level restore.
- unified governance: : Manage access control and auditing through Unity Catalog for a single security model across your entire data estate.
- sync tables: : Keep your operational data and historical lakehouse context in sync without maintaining fragile pipelines.
Get started with Databricks Lakebase today
Building AI Agents and Apps with Lakebase
With Lakebase, operational workloads run directly on the Databricks platform. Applications share the same governance, security, and data foundation that are already relied upon for analytics and AI. There are no secret databases to manage, no separate access controls to maintain, and no data movements to keep in sync.
This shared foundation enables common application patterns such as:
- real time feature service For machine learning with low-latency access to fresh data.
- Permanent memory for AI agents Which is built in line with the lake house.
- Embedded Analytics Which combines operational data with historical insights.
Hafnia used lakebass To move beyond classic BI stacks and static reports toward real-time business applications for fleet, commercial and finance workflows. By using Lakebase as the transaction engine for their internal operations portal, they reduced the time to deliver production-ready apps and dashboards from 2 months to just 5 days.
Other companies like Warner Music Group have enjoyed growth efficiencies thanks to Lakebase.
Lakebase provides us with a unified foundation where analytics and operational workloads work together in real time. By moving insights directly into production systems, we can respond faster, innovate with confidence and ship new capabilities without compromising reliability. This speed and integration is critical as we scale experiences for our customers. – Mike Jones, Director of Software Engineering, Warner Music Group
Enterprise Administration and Performance
The GA release of Lakebase adds production-grade features for reliability, performance, and administration.
- Unified Governance with Unity List: The application inherits the persistent access controls, auditing and compliance on the Databricks platform.
- Trustworthy basis for AI: Governed, auditable operational data ensures that autonomous AI systems act on reliable, compliant information.
- Automatic backup and point-in-time recovery Enables teams to restore database state to a specific millisecond within a configurable retention window, protecting against application bugs or accidental deletion.
- increased storage capacity Supports up to 8TB per instance, enabling larger application workloads.
- Postgres 17 support Continuing to support Postgres 16 brings the latest Postgres improvements and extensions, including pgvector for AI-powered search.
Together, these capabilities make LakeBase suitable for mission-critical systems with high reliability and performance requirements.
Modernize legacy systems with Lakebase
Many organizations run critical applications on legacy databases and desktop tools that are difficult to develop. Changing them often requires rebuilding the entire stack or running parallel systems for years.
Lakebase provides a simple, modernization path by consolidating application logic, analytics, and governance on a single platform.
easyJet used Lakebase and Databricks apps to replace a decade-old desktop application and one of Europe’s largest legacy SQL Server environments. They consolidated over 100 Git repositories into 2 and reduced the development cycle from 9 months to 4.
We’ve left our old systems behind, but with the Databricks Data Intelligence Platform—specifically Lakebase and Databricks Apps—we’ve created a revenue management app that’s faster, simpler, and far more reliable. What once took months of effort is now accomplished in a fraction of the time, laying the foundation for commercial decision making and empowering our teams to analyze, decide and act like a data-driven modern airline retailer. – Denis Michon, Head of Data Products, easyJet
try it today
With general availability, Lakebase establishes a new foundation for operating systems in an AI-native world. Starting today, Lakebase is available in beta for production workloads in select AWS regions and on Azure in select regions. We’re rapidly expanding GA to Azure over the next few months and to Google Cloud later this year. Compliance certifications including SOC2 and HIPAA are also on the roadmap for early this year.
Get up and running with LakeBase in minutes. Create your first project, connect to your database and explore the key features. To get started, check out our technical documentation and our Getting Started Guide.
Lakebase GA comes with validated integrations from partners who have worked closely with Databricks during development and validated Lakebase in real production environments. Read more about our GA partner ecosystem in our partner blog Here.
