CORTEXFLOW · FREE FOREVER

AI Agents

Talk to Your Data: Sheets Chat & Digests

↓ Download workflow JSON

Import in n8n: Workflows → ⋯ → Import from file, then add your credentials. Free download with a free account.

How it works

  1. Chat Trigger — receives your question from the n8n chat panel.
  2. Read Sheet Rows — pulls every row from your Google Sheet deterministically (no AI guessing here).
  3. Format Rows as Table — converts the rows into a compact text table the model can read.
  4. AI Agent — answers your question using only the table data, with gpt-4o-mini under the hood. If the answer is not in your data, it says so instead of inventing numbers.
  5. Digest Schedule — a second, unconnected chain that fires every morning at 8:00.
  6. Read Sheet (Digest) — pulls the sheet rows again for the digest run.
  7. Summarize Changes — GPT reads the data and writes a short bullet-point digest of key figures and changes.
  8. Send Digest Email — emails the digest to you via Gmail.

Set up steps

You’ll need:

  • n8n (self-hosted or cloud)
  • A Google account with access to the spreadsheet you want to query
  • An OpenAI account — get an API key at platform.openai.com → API keys
  • A Gmail account to receive the digest
  1. Download the workflow JSON using the download button on this page.
  2. In n8n, go to Workflows → ⋯ → Import from file and select the JSON.
  3. Create your Google Sheets credential — in n8n go to Credentials → New → Google Sheets OAuth2 API, sign in with Google, and save. Then open Read Sheet Rows and Read Sheet (Digest), pick your spreadsheet in the Document field and the tab in the Sheet field, and select your credential in both nodes.
  4. Create your OpenAI credential — in n8n go to Credentials → New → OpenAI, paste your API key, save. Then open OpenAI Chat Model and Summarize Changes and select your credential in both.
  5. Create your Gmail credential — in n8n go to Credentials → New → Gmail OAuth2 API, sign in, save. Open Send Digest Email, select your credential, and replace [email protected] with your own address.
  6. Activate the workflow (toggle, top-right), then open the chat panel and ask a question about your sheet.

Customize it

  • Change the digest time — open Digest Schedule and edit the cron expression (e.g. 0 18 * * * for 6:00 PM).
  • Change the model — open OpenAI Chat Model or Summarize Changes and pick another model.
  • Add more context — chain extra Set nodes before the agent to pull in more sheets or documents.
  • Digest to Slack instead — swap Send Digest Email for a Slack node.

Workflow nodes

This workflow is built from 9 nodes. Here's what each one does:

#NodeTypeWhat it does
1 Chat Trigger Chat Trigger Starts the workflow from the n8n chat interface.
2 Read Sheet Rows Google Sheets Reads from or writes to a Google Sheet — this workflow's log and state store.
3 Format Rows as Table Set Holds editable configuration values — fill in your own settings here before activating.
4 AI Agent AI Agent The AI brain — reasons over the task and calls tools to get it done.
5 OpenAI Chat Model OpenAI Chat Model The language model powering the AI agent.
6 Digest Schedule Schedule Trigger Starts the workflow automatically on a schedule.
7 Read Sheet (Digest) Google Sheets Reads from or writes to a Google Sheet — this workflow's log and state store.
8 Summarize Changes OpenAI Sends a prompt to OpenAI and returns the AI's answer.
9 Send Digest Email Gmail Sends or reads email through your Gmail account.

Setup instructions

1. Add your credentials

In n8n go to Credentials → New, create each credential below, then open the named node and select it:

NodeCredential to add
Read Sheet Rows Google Sheets OAuth2
OpenAI Chat Model OpenAI API
Read Sheet (Digest) Google Sheets OAuth2
Summarize Changes OpenAI API
Send Digest Email Gmail OAuth2

How this workflow starts

  • Chat Trigger: Fires automatically on its event — activate the workflow and connect the service.
  • Schedule Trigger: Runs every cron.

Use cases

Most small businesses run on a Google Sheet long before they can afford a dashboard. The problem is that getting an answer out of that sheet means filters, pivot tables, and formulas, so questions go unasked and the data quietly goes stale. This template gives you two ways into your data. In chat mode, you ask a question in plain English in the n8n chat panel and an AI Agent answers using only the rows in your sheet. Read Sheet Rows pulls every row deterministically, Format Rows as Table compacts it into text the model can read, and if the answer is not in your data the agent says so instead of inventing numbers. In digest mode, a separate chain fires every morning at 8:00: Digest Schedule triggers, Read Sheet (Digest) pulls the rows again, Summarize Changes has GPT write a short bullet-point digest of key figures and changes, and Send Digest Email sends it to you via Gmail. It is built for owners and managers who live in spreadsheets but would rather talk to them than wrestle them.

The performance marketer watching twelve client accounts

A four-person performance marketing agency keeps one master sheet: client, spend, ROAS, cost per lead, updated daily. Monday mornings used to mean an hour of filtering before the 10 AM client calls. Now the 8:00 AM digest lands in the founder's inbox with the key figures and what moved overnight. When a client emails asking why spend spiked, the account manager opens the chat panel and asks which clients are over their weekly budget. Read Sheet Rows and Format Rows as Table feed the live data to the AI Agent, so the answer reflects today's numbers, not last week's memory.

  • Variation: shift the Digest Schedule cron to 6 PM for an end-of-day recap the team reads before stand-down, or swap Send Digest Email for a Slack node so the digest lands in the client-ops channel where everyone sees it.

The Airbnb host with fifteen properties

A host with fifteen listings tracks occupancy, nightly rate, and cleaning status in a single sheet, and the quiet killer is a property sitting empty for three weeks unnoticed. The morning digest flags occupancy drops and units with no upcoming bookings, so problems surface before the month's revenue does. Between digests, the chat answers the questions that used to require scrolling: which properties are below 50 percent occupancy next month, which units still show cleaning as pending.

  • Variation: keep a second tab for cleaner schedules and chain an extra Set node before the agent so one question can join occupancy with cleaning status.

The D2C ops manager fighting stockouts

A direct-to-consumer brand keeps 200 SKUs in a sheet with current stock and reorder points. Running out of the bestseller for a week costs more than any ad budget can fix. The daily digest lists every SKU below its reorder level plus the biggest week-over-week stock drops, so the purchase order goes out before the shelf is empty. For ad-hoc checks the ops manager asks which SKUs fell more than 20 percent this week, or how many units of the top ten sellers remain.

  • Variation: pull supplier lead times from a second sheet with an extra Set node before the agent, and ask which SKUs will run out before their next delivery arrives.

The gym owner chasing late payments

A 300-member gym tracks memberships and payment status in a sheet, and around forty members are usually overdue at any time. The digest arrives Monday morning with the newly overdue accounts, so the front desk knows exactly who to call that week. In chat, the owner asks who owes more than two months, sorted by amount owed, and gets a clean list instead of a filtered view they have to build themselves.

  • Variation: move the Digest Schedule to Monday at 9 AM so it lands right as the front-desk shift starts the week's collection calls.

The freelance bookkeeper with five clients

A freelance bookkeeper manages invoicing for five small businesses, each with its own invoice sheet, and the context switching is the real tax. With one workflow per client sheet, the bookkeeper opens the chat and asks for total overdue and the three oldest unpaid invoices, per client, in under a minute. The morning digest per sheet flags newly overdue invoices, so nothing ages past thirty days unnoticed.

  • Variation: change the digest time per client to match their billing rhythm, or switch the OpenAI Chat Model to a stronger model for sheets with messier, less structured rows.

← All templates