Notion

The Notion integration lets your bot read from your databases — docs, knowledge, product data. Connect with Notion, choose which databases and properties to expose, and the bot answers from them.

Connecting#

  1. Open the Notion tab

    Go to Integrations → Notion.
  2. Authorize with Notion

    Sign in via OAuth and grant access to the workspace. You control which pages and databases the integration can reach.
  3. Select databases

    Choose the databases the bot should read from.

Mapping properties#

For each connected database, decide what the bot sees:

  • Visible columns — pick which properties to expose and keep the rest private.
  • Field mapping — map Notion properties so the AI understands what each one means.

Name properties clearly

Descriptive property names help the bot reason about your data. A column called Status is easier to use than one called Col 3.

Linking databases#

You can link one database to another so a single AI action updates both at once — for example an Orders database linked to an Inventory database, so every new order also decreases stock. The linked resource can be another Notion database or a Google Sheet.

  1. Pick a linked resource

    In the database settings, open Linked resource (optional) and choose what to link.
  2. Auto-link with AI

    Describe the rule in plain words — e.g. “when a new order is added, decrease product stock by quantity” — and click Auto-link. The AI proposes the property mapping for you to confirm or adjust.
  3. Match by & Apply amount

    Set Match by (the property that joins the two, e.g. product name or SKU) and Apply amount (the number to change), then pick + or for the operation.

One action, two databases

Once linked, the bot exposes a single composite action that finds the matching entry in the linked database and applies the change there automatically.

Test & attach#

Test the connection to confirm the bot reads the right entries, then attach the integration to a bot from its Tools & integrations settings. Updates in Notion are reflected on the next read.