Google Cloud brings conversational analytics to Gemini with data connectors

Natural-language access to live enterprise data

Data Cloud connectors give business users and analysts a way to ask questions about enterprise data directly from their workspace, according to Google Cloud. The company says users can move results into everyday tools, including a Google Doc, an email draft in Gmail or a presentation in Google Slides.

Gemini can connect to BigQuery in preview, operational databases such as Spanner, AlloyDB and Cloud SQL, and governed data from Looker, also in preview. Users do not need to know which table or project contains the relevant information. For example, a question about sales of cold weather products can be mapped to database categories such as sweaters, coats and jackets.

The feature is built around Knowledge Catalog. It searches for relevant data, interprets business terms using technical metadata, business glossaries and semantic models, then generates and runs a SQL query through the connector.

Permissions stay tied to the user

Google Cloud says connector queries execute under the end user’s credentials and respect existing row- and column-level permissions. An employee who cannot access executive salary data or protected healthcare records cannot retrieve that information through a natural-language prompt.

Knowledge Catalog only exposes metadata that the user is allowed to access. It also provides a context bundle containing technical metadata, business logic, table relationships and verified queries. According to Google Cloud, this context is intended to improve query accuracy while reducing latency and token consumption.

The BigQuery connector also extends conversational analytics to what Google Cloud calls a borderless lakehouse. It can query open table formats, including Apache Iceberg, Delta Lake and Parquet, as well as cross-cloud datasets in AWS or Azure through BigQuery. The connector inherits fine-grained access policies from BigQuery and Knowledge Catalog.

Connectors and data agents serve different workloads

Google Cloud distinguishes direct connectors from Conversational Analytics agents. Connectors are intended for exploratory and ad hoc analysis, while data teams can build and certify scoped agents for governed, multi-step strategic analysis. These agents can be created across BigQuery, Looker, Spanner, AlloyDB and Cloud SQL, then published in Gemini or made available through an API for enterprise applications.

Administrators configure connectors in the Gemini admin console, creating default read-only connections. The system uses federated query mode, so data is queried in place rather than copied or indexed into a Gemini data store. Google Cloud also says first-party connectors offer engine-specific skills, guided setup and managed authentication compared with public or self-managed MCP servers.

For users, the announcement means Gemini can become a direct interface for governed enterprise data without requiring analysts to handle every request manually. For data teams, the combination of permission-aware connectors and certified agents offers separate paths for flexible exploration and controlled reporting, according to Google Cloud.