# How to connect BigQuery to an AI agent: what GTM AI agents can do with BigQuery in Day AI

> How to connect BigQuery to an AI agent: in Day AI, a workspace Owner saves your Google Cloud OAuth app once, each person signs in with Google, and GTM AI agents explore datasets and run SQL next to every customer conversation.

Updated November 20, 2026

**The short answer.** To connect BigQuery to an AI agent, register an OAuth app in your own Google Cloud project for Google's BigQuery MCP server and have each person sign in to Google, so [GTM AI agents](/guides/gtm-ai-agents), [AI for GTM](/guides/ai-for-gtm) pointed at your warehouse, can answer warehouse questions in the same place they read every customer conversation. In Day AI, a workspace Owner saves that Google Cloud app once under Settings > Connectors, each person connects their own Google account, and their agent can explore datasets and table schemas and run SQL against your Google Cloud projects, whether you use Day AI in place of your CRM or on top of Salesforce or HubSpot.

![The BigQuery connector in Day AI's Connectors settings](/guides/images/connect-bigquery-to-day-ai/connector.webp)

Imagine you run RevOps at a logistics-software company, and at 4:30 on a Wednesday your CRO asks a simple question: which customers who raised pricing on a call this quarter also cut their shipment volume? The pricing half lives in conversations, and the volume half lives in a BigQuery table you've queried maybe twice. You ask your Day AI agent. It finds the accounts where pricing came up in meetings and email, looks through the BigQuery dataset to find the table with shipment volume, writes the SQL, and runs it. It comes back with eleven accounts, each with the call where pricing came up and the volume trend next to it, plus the query it ran so you can check the logic. You ask it to drop two accounts that are mid-migration, then to put the list on a page for the Thursday pipeline review. The list was on the page before Thursday's agenda went out.

**Where the usual answer goes wrong.** Guides to AI on BigQuery mostly stop at turning a question into SQL, and that part works. The trouble is that the most important go-to-market questions are half warehouse and half conversation, and the warehouse doesn't know what a customer said on Tuesday. Day AI treats the conversations as the record and BigQuery as the place the numbers live, and the agent reads both before it answers. If you already have a text-to-SQL tool or a data team that turns around requests, keep it. What changes is that the answers now sit next to the customer's own words.

## The ask, query, check method for BigQuery, with Day AI

A warehouse answer is only as good as the question and the logic behind it. Three moves keep both honest.

1. **Ask in the business's words.** Reps and leaders ask about accounts, renewals and pricing, not tables. In Day AI, you ask the agent in plain words, and it reads the relevant customer conversations before it goes to the warehouse.
2. **Query the right table.** Half the work of a warehouse question is finding where the data lives. In Day AI, the agent explores BigQuery datasets and table schemas, writes the SQL and runs it against your Google Cloud projects, within your own BigQuery access.
3. **Check against what customers said.** A number that disagrees with the last three calls deserves a second look. In Day AI, the agent shows the query it ran and the conversations next to the result, so you can test the answer before you share it.

## How to connect BigQuery to Day AI

BigQuery is in beta in Day AI. Setup has two parts: one for a workspace Owner, then one for each person.

### Step 1: Create an OAuth client in your Google Cloud project for Day AI

In the Google Cloud project your company uses for BigQuery, create an OAuth client for Day AI, and add the redirect URL shown in Day AI's BigQuery setup, `https://day.ai/mcp-oauth-redirect`. Copy the client ID and client secret. Your Google Cloud admin is the right person for this step.

### Step 2: Save the app in Day AI (workspace Owner)

In Day AI, open Settings > Connectors, find BigQuery, and choose **Set up for workspace**. Paste the client ID and client secret (both are required) and save. Until then, the connector shows **Needs workspace setup**, and no one in the workspace can connect BigQuery. Saving the setup doesn't connect anyone; it lets them connect.

### Step 3: Each person connects their Google account in Day AI

Each person opens Settings > Connectors, finds BigQuery, clicks **Connect**, and signs in with their Google account. Their agent then works with that person's own BigQuery permissions, so an analyst and a rep can connect the same connector and see different datasets.

### Step 4: Set BigQuery's tool permissions in Day AI

On the BigQuery connector's page, each tool under **Tool permissions** can be Always allow, Needs approval or Off. Exploring schemas doesn't interrupt the chat, and any BigQuery tool not marked read-only asks first with an approval card; if running a query asks first and you want it to run straight through, set that tool to Always allow.

### Step 5: Ask your Day AI agent warehouse questions with context

Try "Which accounts renewing next quarter have lower usage than at signing, and what did they say on their last call?" or "Show me the schema for our billing dataset, then total last month's expansion revenue by segment." Ask the agent to show its SQL whenever you want to check the logic. For a recurring report, ask it to create a skill and pick BigQuery in the skill's **Connectors** picker; the skill runs with its owner's connection. BigQuery sits alongside [every Day AI integration and connector](/integrations), so the same question can reach your CRM data too.

## Where a data request queue breaks

- **Every question waits in line.** A ticket to the data team takes days for a question asked before a Friday review. In Day AI, the person asking runs it with the agent in the same chat.
- **The query can't see the conversation.** A warehouse join can't include what a customer said on a call. In Day AI, the agent reads meetings and email next to the query result.
- **Shared credentials see too much.** A service account for an AI tool gives everyone the same reach. In Day AI, each person signs in with their own Google account, so the agent sees what they see.
- **Nobody can check the logic.** A number in a chat with no query behind it is hard to trust. In Day AI, the agent shows the SQL it ran on request, and you can ask it to change the logic.

## Frequently asked questions

### How do I connect BigQuery to an AI agent?

Use Google's BigQuery MCP server with an OAuth client from your own Google Cloud project, and have each person sign in with their own Google account. In Day AI, a workspace Owner saves that OAuth client's ID and secret under the BigQuery connector's Set up for workspace, and then each person clicks Connect under Settings > Connectors.

### What's the best AI agent for querying BigQuery about customers?

Day AI answers go-to-market questions on BigQuery in one reply, because its agent runs the query and reads the account's meetings, email and calls together. Each person signs in with their own Google account, so the agent sees only the datasets that person can see.

### Can Day AI write SQL against our BigQuery data?

Yes. Your Day AI agent can explore BigQuery datasets and table schemas, write the SQL for a question you ask in plain words, and run it against your Google Cloud projects, inside your own BigQuery access.

### Who sets up BigQuery in Day AI?

A workspace Owner (or anyone with permission to manage workspace settings) saves the Google Cloud OAuth client once. After that, every person connects their own Google account; the saved app is shared, but nobody shares a login.

### Does BigQuery data get copied into Day AI?

Not at all. Your agent runs the query live when you ask, so the result reflects your warehouse as it is right now, and it never reaches past what your own Google account can query.

### What is Day AI?

Day AI is the customer memory and agent platform for go-to-market teams. Even the best CRM only holds a fraction of what actually happened with your customers. Day AI is what comes after CRM. It builds a self-updating customer record from every meeting, email, and call, and runs governed agents that create finished work on top of it, with every fact and every action traced to source conversations. It does everything teams use a CRM for, without the data entry, and does what a CRM never could. Agents prospect, prep every meeting, draft follow-ups, keep the pipeline clean, and much more, giving every rep added selling capacity that used to require hiring additional people. It runs in place of Salesforce or HubSpot or as a layer on top of them.

Canonical: https://day.ai/guides/connect-bigquery-to-day-ai
