Google BigQuery

Connect AI agents to Google BigQuery. Agents run GoogleSQL with named parameters, stream rows into tables, read table data without query cost, and manage datasets and tables.

11 actions available

A scheduled flow needs a daily KPI computed in the warehouse

Agent runs Run Query with a named parameter to compute the KPI

If the job outruns the timeout, agent polls Get Query Results by job id

Agent reads the rows as plain JSON keyed by column name

Agent posts the numbers to the metrics channel via Slack

Agent runs Insert Rows to stream events from another service into a table

Any Delete Dataset or Delete Table pauses for approval

What This Integration Enables

Agents run analytical SQL and use the resulting rows directly in flow logic, parameterize queries safely with `@name` placeholders, stream events into a table in near real time, page through large or long-running results by job id, and provision datasets and tables. Run Query waits up to a timeout and, if the job has not finished, returns a job id that Get Query Results uses to fetch the rows once the job completes. Streaming inserts are queryable within seconds. Delete Dataset (with delete-contents) and Delete Table permanently remove data.

Without FlowRunner

Insights stuck in dashboards KPIs live in BI tools and never drive an action

Streaming needs code Loading events into the warehouse means a custom pipeline

Destructive drops unguarded A Delete Dataset or Delete Table removes data with no checkpoint

With FlowRunner

Warehouse drives the flow Query results feed flow logic and post to channels directly

Streaming as a step Insert Rows streams events into a table without a custom pipeline

Guarded drops Delete Dataset and Delete Table pause for human approval

Use Case Scenarios

KPIs to a channel

On a schedule, the agent runs Run Query to compute the day's KPIs, reads the rows, and posts them to a metrics channel with Slack. Numbers that used to sit in a dashboard now show up where the team already works, and the flow can branch on them.

Streaming events into the warehouse

Another connector produces records, for example payment intents on a schedule. The agent calls Insert Rows to stream each record into an analytics table, so warehouse-side reporting stays current without a bespoke ingestion pipeline.

Cleanup with a human gate

An agent is asked to remove an obsolete dataset or table. Before it runs Delete Dataset or Delete Table, it does not act on its own. It routes the target and a summary for approval, and drops it only after a person confirms.

Human-in-Loop Highlight

Reading and streaming into BigQuery are safe; deleting a dataset or a table is the action that cannot be undone from the flow. FlowRunner's answer is human-in-the-loop, an execution pattern where the agent pauses on its own, assembles the context and the choices, routes to a human on their preferred channel, and resumes the moment they respond. When a flow reaches Delete Dataset (with delete-contents enabled) or Delete Table, the agent pauses before executing and asks through Slack: "Ready to Delete Table `analytics.payments_2024`. Here is its row count and last-modified time. Approve or cancel?" The delete runs only after a person confirms, with the approver and timestamp captured in the run log. A connector can drop any table; an orchestration layer knows which drops should stop and ask.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via Slack

Human decides

Agent resumes with decision

Agent Capabilities

11 actions

Queries

Run Query Runs a GoogleSQL query with named `@name` parameters and returns rows as plain JSON keyed by column name. If the job outruns the timeout, it returns a job id for Get Query Results to page through.
Get Query Results Fetches the rows of a completed job by id and pages through large result sets with a page token.

Table Data

Insert Rows Streams rows into a table; inserted rows are queryable within seconds. Partial failures are reported per row.
List Rows Reads a table's contents without incurring query cost, for exports and syncs.

Datasets

List Datasets Lists the datasets in the project, for discovery.
Create Dataset Creates a dataset in the configured location.
Delete Dataset Permanently removes a dataset. With delete-contents enabled, its tables go with it. This cannot be undone.

Tables

List Tables Lists the tables in a dataset.
Get Table Returns a table's metadata and schema.
Create Table Creates a table, including nested Record columns.
Delete Table Permanently removes a table and its data. This cannot be undone.

Start building with Google BigQuery

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us