AI News

Google’s New AI Agents Take the Grunt Work Out of Database Ops

Quick answer

Google Cloud's new AI agents automate database onboarding and observability, cutting manual work and speeding up troubleshooting. Here's the scoop.

Google Cloud just dropped two new AI-powered database agents at Next ’26, and honestly, they’re the kind of help every developer dreams of. Say goodbye to manual schema tweaking and hello to a virtual DBA that’s always on duty.

These agents are part of the Agentic Data Cloud launch, and they’re designed to handle the boring, repetitive parts of database management so you can focus on building cool stuff.

Meet the Database Onboarding Agent

Remember the days of reading endless docs to pick the right database? The Onboarding Agent is here to change that. Just describe your app’s needs in plain English—IOPS, latency, replication lag—and it’ll recommend the perfect Google Cloud database for the job.

It’s like having a capybara guide you through the swamp, pointing out the safest channels and the best basking spots. No more guessing games.

  • Smart recommendations: It reflects your requirements back, so you know why it suggested AlloyDB for high availability.
  • Streamlined setup: Once you pick a service, it generates the commands to provision and configure everything.

Database Observability Agent: Your 24/7 Watchdog

This agent is all about Day 1 and Day 2 operations. It monitors, troubleshoots, and even fixes issues—with your approval, of course. It’s like having a caiman on your team, but one that’s on your side.

It connects telemetry from Database Insights, Cloud Monitoring, Cloud Logging, and Cloud Trace to give you root cause analysis in minutes. No more hours of digging through logs.

  • Fleet-level troubleshooting: Ask questions like “Which databases consumed the most CPU?” and get a summary across your entire fleet.
  • In-product investigations: Correlates complex telemetry to pinpoint latency spikes or lock contention.
  • Validated remediations: Suggests actions like adding indexes, with rationale and expected impact.
  • MCP tools: Available via Database Insights and Database Center MCP servers, so you can integrate them into your IDE.

Integration That Fits Your Workflow

You can chat with these agents in the Google Cloud console, via CLI, or right in your IDE. They’re also available through Gemini Cloud Assist and MCP servers, so you get help where you need it.

It’s like having a virtual DBA in your pocket, ready to answer questions and fix issues before they become disasters.

Supported Databases

The Observability Agent works with a bunch of Google Cloud databases, including:

Whether you’re dealing with query hotspots, lock contention, or replica lag, these agents have your back.

Get Started

Both agents are available now via Gemini Cloud Assist. Check out the docs for AlloyDB, Cloud SQL, and Spanner to see them in action.

If you’re comparing database platforms, don’t forget to check out our reviews on Supabase, Neon, and Firebase to see how they stack up against Google’s offerings.

Original announcement published on Google Cloud.