Lakebase
#5 in Enterprise Agent PlatformsDatabricks · 2× · last seen Jul 19, 2026
56
Momentum
Lakebase is a fully managed, serverless Postgres database from Databricks integrated as an operational (OLTP) database layer into the Data Intelligence Platform. It is built on Neon technology, separates compute and storage, and serves as a persistent state store for AI agents as well as a backend for data-driven applications with native Unity Catalog governance. Key features include autoscaling including scale-to-zero, instant branching/cloning, point-in-time recovery, and change-data-feed synchronization with the lakehouse. The product was introduced as a Public Preview in June 2025 and is now generally available (GA) on AWS and Azure, with GCP support announced.
Momentum trend
20.04.19.07.
Features
| Compliance/Certification | GA for workspaces with compliance security profile under HIPAA, C5, TISAX, or None standards; embedded in Databricks platform with SOC 2 Type II, ISO certifications |
| Deployment Model | Fully managed serverless SaaS with decoupled compute/storage, autoscaling and scale-to-zero; deployed as a database instance/project within a Databricks workspace |
| Use Case Scope | Operational database for AI agents (persistent state/memory), transactional applications, feature store for ML models, real-time analytics on operational data |
| Integrations | Unity Catalog, Databricks Apps, Lakehouse Federation, Change Data Feed/Lakehouse Sync to Delta tables, Mosaic AI Model Serving, standard Postgres drivers/psql, extensions such as pgvector and PostGIS |
| License | Proprietary managed cloud service (no open-source license model); built on open-source Postgres engine |
| Platform | Databricks Data Intelligence Platform (Lakehouse); available on AWS and Azure, GCP support not currently available |
| Price | Pay-as-you-go with 14-day free trial; Operational DB (Lakebase) compute starting around $0.069/CU; Always-On pricing offers 25% lower rate on baseline capacity; custom committed-use discounts available |
| Release Date | Public Preview: June 11, 2025; GA on AWS: February 3, 2026; Beta/GA on Azure from March 2026 |
