Gemini Enterprise Federated Query for Data Cloud Connectors
On 2 October 2026 Google made federated query mode (Preview) available for Gemini Enterprise Data Cloud connectors: querying operational and analytical data in place, using the Model Context Protocol (MCP) and each user's own credentials, without copying or ingesting data into a data store.
Tool Overview
Features, steps and FAQ below
Features
- ✓ Google says federated query mode (Preview) is available for the AlloyDB for PostgreSQL, BigQuery, Cloud SQL and Spanner connectors
- ✓ Google says the mode queries data in place using the Model Context Protocol (MCP) and each user's own credentials, without copying or ingesting data into a data store
- ✓ Google says that when you connect a BigQuery, Spanner, Cloud SQL or AlloyDB for PostgreSQL data store in federated query mode, Knowledge Catalog (Preview) is automatically enabled on the Assistant tab of your Gemini Enterprise app
- ✓ Google says use cases include querying and analyzing structured data in BigQuery with natural-language prompts, and running live analytical queries and authorized database operations across Spanner, Cloud SQL and AlloyDB for PostgreSQL
- ✓ Google says Data Cloud connectors have two modes: in data ingestion mode they copy data from the source into the Gemini Enterprise data store, while in federated query mode they query data in place without copying or indexing it
How to Use
- In Gemini Enterprise, configure a Data Cloud connector for BigQuery, Spanner, Cloud SQL or AlloyDB for PostgreSQL
- Choose federated query mode (Preview) to query data in place with each user's own credentials, without copying or ingesting it
- Use the automatically enabled Knowledge Catalog tools on the Assistant tab to search the data users can access and look up context
- Query structured data with natural-language prompts, or run live analytical queries and authorized operations against operational databases
FAQ
What is federated query mode?
Google says federated query mode is a Preview mode for Gemini Enterprise Data Cloud connectors that queries operational and analytical data in place using the Model Context Protocol (MCP) and each user's own credentials, without copying or ingesting data into a data store.
Which data sources are supported?
Google says federated query mode (Preview) is available for the AlloyDB for PostgreSQL, BigQuery, Cloud SQL and Spanner connectors.
How is it different from data ingestion mode?
Google says in data ingestion mode, Data Cloud connectors copy data from the source into the Gemini Enterprise data store, whereas federated query mode queries data in place without copying or indexing it.
How does Knowledge Catalog fit in?
Google says that when you connect a BigQuery, Spanner, Cloud SQL or AlloyDB for PostgreSQL data store in federated query mode, Knowledge Catalog (Preview) is automatically enabled on the Assistant tab of your Gemini Enterprise app, giving the agent tools to search data the user can access and look up context.
What are typical use cases?
Google says use cases include querying and analyzing structured data in BigQuery with natural-language prompts, and running live analytical queries and authorized database operations across Spanner, Cloud SQL and AlloyDB for PostgreSQL.