CoreWeave has announced the general availability of CoreWeave Database, a managed PostgreSQL service that operates within customers’ CoreWeave Kubernetes Service (CKS) clusters.
The new offering enables AI teams to use a production-grade relational database alongside their graphics processing unit (GPU)-powered training and inference workloads.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
The introduction of CoreWeave Database means users can now deploy, manage, and scale production applications on a single platform, removing the need for separate database management or cross-cloud setups.
Previously, those building on CoreWeave either managed PostgreSQL themselves or relied on external cloud providers, a process that added network latency, egress paths, and extra security considerations.
This challenge has increased as AI projects move towards production, with applications, training jobs, and retrieval-augmented generation (RAG) services frequently accessing databases for state management, metadata, and embeddings.
CoreWeave Database operates entirely within customers’ CKS clusters, running databases on CPU nodes and connecting applications through in-cluster Kubernetes services. This approach eliminates the need for network traffic across cloud boundaries or managing additional egress paths.
High availability is provided, with each database maintaining a primary instance and read replicas spread across hosts, along with automatic failover.
Data storage uses CoreWeave Distributed File Storage (DFS), and continuous backups offer point-in-time recovery within a retention window.
Customers retain control of their data, schemas, queries, and users, while CoreWeave manages replication, failover, patching, and backups.
CoreWeave product senior vice president Corey Sanders said: “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.
“CoreWeave Database puts that data in the same cluster as the compute that depends on it, operated by CoreWeave so teams can spend their time on models instead of database operations.”
The database service is said to be fully compatible with PostgreSQL, allowing migration from other clouds or self-managed deployments using native Postgres tools.
Security features include private, cluster-only access, encryption at rest and in transit, managed TLS certificates, and granular control through PostgreSQL roles and Kubernetes network policies.
CoreWeave Database leverages CloudNativePG, an open-source Kubernetes operator for PostgreSQL, for management of primary and replica instances as well as failover.
The company contributes to the wider PostgreSQL ecosystem, including recent support for S3 compatibility improvements in Barman and configuration enhancements in pgexporter.
This release marks an expansion of CoreWeave’s platform from compute towards broader cloud services for AI application deployment.
Last month, CoreWeave introduced the CoreWeave Zero Trust AI Framework, which provides security controls intended to help customers manage access to their AI models, data, and runtime environments.
