Storage Intelligence advisor: Know what changed in your storage estate and act on it
Google Cloud Storage Intelligence advisor reaches general availability, providing automated insights into storage changes without custom pipelines. The feature uses curated findings to identify anomalies in access patterns and metadata, helping teams avoid unexpected costs. It works alongside expanded batch operations for executing decisions across millions of objects. No new underlying storage primitives are introduced, only a higher-level management interface.
Google Cloud announced that its Storage Intelligence advisor has reached general availability, accompanied by expanded batch‑operation capabilities. The advisor automatically scans a project’s buckets, using daily snapshots of usage and metadata to surface anomalies such as spikes in Class A or B operations, 429 errors, or cross‑region egress. Findings appear in the Cloud Console without the need to build pipelines, dashboards, or schemas, and customers can enable the service at the organization, folder or project level. The feature gives teams immediate visibility into changes across storage estates that can otherwise hide for weeks on an invoice, helping avoid unexpected retrieval fees from cold‑storage access. By coupling the advisor with serverless batch operations, recommendations can be turned into policy‑driven jobs that act across millions of objects, reducing manual effort and throttling concerns. Adoption appears strong: the number of customers analyzing over 1 billion objects with Storage Intelligence has more than doubled this year, and in the last 30 days more than 6 000 findings were generated across hundreds of users. It remains unclear how the advisor will perform in environments with highly volatile access patterns that shift from week to week, as the baseline relies on recent activity. Google has not disclosed any impact on underlying storage primitives, only that the management interface is new. The long‑term cost‑benefit balance for organizations that already run custom pipelines is also uncertain.