Artificial Intelligence

CoreWeave bets AI apps need Postgres nearby

CoreWeave launched a managed Postgres inside CKS to keep AI state next to GPUs.

StartupHub.ai Staff

By StartupHub.ai Staff, reviewed by Daniel Singer, Editor

2 min read

CoreWeave Database runs Postgres inside GPU clusters
Credit: CoreWeave Newsroom

CoreWeave is right: production AI breaks when the database lives in another cloud.

CoreWeave made that case with general availability of CoreWeave Database, a managed PostgreSQL service that runs inside customers’ CoreWeave Kubernetes Service clusters, the company said.

I think this is the least glamorous and most necessary step in its platform push.

Agents read and write state between tool calls. Training jobs log metadata. RAG services query embeddings. When those hits cross clouds, latency and ops cost compound, and CoreWeave wants that traffic to stay in-cluster on CPU nodes with a Kubernetes service and no egress path, according to the announcement.

That fits how CoreWeave describes its platform, as Forge for training, agents and inference paired with Infrastructure for GPU compute, storage and networking. Database is the state layer for that loop, not a side product.

The two details that matter are high availability and data protection done by CoreWeave. Each database runs a primary with replicas across hosts, with automatic failover on Distributed File Storage, plus continuous backups and point-in-time recovery to any moment in the window, the company said.

Corey Sanders, SVP of product, whose team owns the pitch to builders, said AI in production runs on state, metadata and embeddings touched on every step and teams should spend time on models instead of database operations.

The concession is fair. Running Postgres on Kubernetes is still operationally hard, and CoreWeave is leaning on CloudNativePG for primary and failover handling while it adds console workflows, backups and monitoring around it.

My read survives that caveat because CoreWeave is also working upstream, with S3 compatibility work in Barman and connection work in pgexporter, plus native pgvector so teams can store embeddings and app data together, the company said. Advaita Bio CEO Mike Mattacola said migration from another cloud was straightforward and his biomedical literature database now runs alongside the apps that use it.

If your agents live on CoreWeave, keep your state there too.

Sources

Get the free daily brief every morning

Every morning the day's AI news, new AI startups and new funding rounds. Free and takes a few minutes to read, unsubscribe at any time.

Join 64,000 other subscribers.

TopicsCoreWeaveCoreWeave DatabasePostgreSQLpgvectorCloudNativePGCoreWeave Kubernetes ServiceAdvaita Bio

StartupHub.ai Staff

Written by

StartupHub.ai StaffEditorial team

The staff writers of StartupHub.ai, ranging from investment analysts to avid AI tool users, early adopters and critical enthusiasts. Backgrounds span engineering, business and the arts. We hold every piece to rigorous standards of research and review.

Daniel Singer

Reviewed by

Daniel Singer is an AI expert who has been writing about, covering and building AI technology since 2018.

Test your Agent Readiness now

Paste your website. How well ChatGPT, Claude and the other agents can read and use a site, in about a minute.

Watch

The videos behind the news, one flip at a time

The feed mixes the videos from our articles with startup demos and you flip through them like TV channels for free with no signup.