CoreWeave Launches Managed PostgreSQL Inside AI Compute Clusters

CoreWeave announced on November 11, 2026 that CoreWeave Database is generally available, providing managed PostgreSQL inside customers’ CoreWeave Kubernetes Service (CKS) clusters alongside AI training and inference workloads.

Databases run on CPU nodes within a customer’s CKS cluster, and applications connect through an in-cluster Kubernetes service. Teams therefore do not need to manage a cross-cloud link or egress path for those database connections, while their relational data remains on the same platform as their GPU workloads.

CoreWeave said customers previously had to administer PostgreSQL themselves or keep it in another cloud, adding a network hop, an egress path and a second environment to secure. Cross-cloud database calls can also increase latency and operational overhead as production applications generate more database activity.

“AI in production runs on more than GPUs. It runs on the application state, metadata, and embeddings that agents and training pipelines touch on every step,” said Corey Sanders, CoreWeave’s senior vice president of product.

Resilience, Security and Vector Retrieval

A primary PostgreSQL instance and its read replicas are distributed across hosts, with live data residing on CoreWeave Distributed File Storage (DFS). CloudNativePG, an open-source Kubernetes operator for PostgreSQL, handles primary and replica management and automatic failover. CoreWeave supplies the surrounding managed service, including monitoring and ongoing operations.

Continuous backup is enabled by default. Point-in-time recovery can restore a database to any moment covered by its retention window, while scheduled and on-demand backups provide additional recovery options. A full backup can also be restored into a new cluster, allowing teams to validate the copy without replacing the original deployment.

Databases remain private to their clusters, with encryption applied at rest and in transit. CoreWeave manages TLS certificates and security patches. Customers control access through native PostgreSQL roles, Kubernetes network policies and audit logging, dividing infrastructure operations from application-level permissions.

CoreWeave says pgvector support allows teams to store embeddings and conduct vector searches for retrieval-augmented generation (RAG) and semantic retrieval in the same database as their application data. That removes the need to operate a separate vector store for those workloads and places retrieval data and application state under one PostgreSQL operational model.

PostgreSQL Migration and Kubernetes Controls

CoreWeave describes the service as fully compatible with PostgreSQL, allowing databases to be moved from another cloud or a self-managed deployment using native Postgres tools. Customers retain ownership of their data, schemas, queries and database users, while CoreWeave handles patching and ongoing database operations.

Management can fit into existing Kubernetes workflows. CoreWeave’s product documentation lists the CoreWeave Cloud Console, kubectl, GitOps tools such as Argo CD or Flux, and custom controllers built against the Kubernetes API as management options.

The resource model includes CWDBCluster, which defines database topology, compute, storage, access and restore settings. A second resource, CWDBBackup, creates on-demand backups and can establish known recovery points before potentially disruptive operations.

Advaita Bio CEO Mike Mattacola said his company migrated its biomedical literature database from another cloud to CoreWeave Database and reached production quickly with hands-on assistance from CoreWeave. He said the managed PostgreSQL deployment now operates alongside the applications that depend on it.

For CoreWeave, the launch adds a managed data-service layer to an infrastructure platform built around accelerated computing. For AI teams, the operational consequence is that application state, vector retrieval and GPU workloads can remain within one cluster environment, with routine database administration handled by CoreWeave and control over data design and access retained by the customer.