Google Cloud’s Conversational Analytics Goes GA Across Data Cloud

Google Cloud is making a big splash in the data swamp, bringing its Conversational Analytics (CA) to general availability across BigQuery and Looker, with previews for AlloyDB, Cloud SQL, and Spanner. Now your capybara builders can chat with their data across multi-cloud and database workloads without breaking a sweat.

Whether your data lounges in Google Cloud or across AWS and Azure, CA agents can query it natively. For data practitioners, it’s baked right into BigQuery Studio, Data Canvas, and Database Studio. Business teams get the same superpowers in Looker, Data Studio, and Gemini Enterprise. And with APIs and MCP tools, you can embed CA into custom apps, Slack bots, or multi-agent orchestrators—wherever your herd roams.

Enterprise-Grade Governance and Cost Controls

Scaling gen AI to thousands of users requires ironclad governance. CA includes Customer Managed Encryption Keys (CMEK), Private IP, VPC controls, and HIPAA compliance. Data residency is guaranteed in the EU and US. Role-based access with parameterized secure views ensures users only see what they’re authorized to—down to row and column level.

Administrators get native cost controls to limit query sizes, plus OpenTelemetry metrics for fleet health, latency, and token consumption. Feedback loops let you review traces and user feedback to continuously improve accuracy.

Co-Designed for Grounded, Deterministic Answers

Wrapping a generic LLM around your database can lead to hallucinated logic—like a caiman mistaking a log for lunch. Google co-designed CA agents alongside the data platforms they query. Agents leverage Knowledge Catalog for data discovery, BigQuery Graphs for multi-hop relationships, and Looker’s semantic layer (LookML) for deterministic metric definitions. No more guessing SQL joins.

Built-in capabilities like ai.search, ai.forecast, and ai.detect_anomalies (powered by TimesFM) let you query multimodal data, forecast trends, and spot anomalies. The ai.key_drivers function performs automated contribution analysis to pinpoint what’s driving unexpected changes.

Proactive Insights with Agentic Workflows

Analytics is moving beyond reactive Q&A to proactive intelligence. With Agentic Workflows (preview), you can schedule automated reporting routines that deliver daily or weekly summaries straight to your chat. Streaming anomaly detection can launch an agent the moment a key metric deviates from baseline—like a caiman sensing a ripple in the water.

Flexible Integration with APIs, SDKs, and MCP

CA is available via native SDKs for Node.js, Java, Go, Python, PHP, Ruby, and .NET. You can publish agents to Gemini Enterprise or integrate them into multi-agent systems using the Agent Development Kit (ADK) and Model Context Protocol (MCP). For example, a supply chain orchestrator agent can query a financial data agent to calculate the margin impact of a shipping delay in real time.

Ready to dive in? Check out the Conversational Analytics documentation and sign up for previews. For more on Google Cloud’s data tools, see our Google Cloud Review.

Original announcement published on Google Cloud.