Build a Slack-based CRM assistant with HubSpot and Google Gemini

How it works

This workflow creates a Slack-based CRM assistant that allows users to query HubSpot data using natural language. When a user mentions the bot in Slack, the message is cleaned and processed to remove Slack-specific formatting. The workflow then retrieves and filters relevant data from HubSpot (deals, companies, and contacts). Finally, an AI agent formats the response and sends a structured reply back to Slack.

Step-by-step

• Trigger on Slack mention

  • Slack Trigger – Listens for app mentions in Slack channels.
  • Code in JavaScript – Cleans the message by removing Slack IDs and formatting.

• Fetch and filter CRM data

  • Get Deals – Retrieves deals from HubSpot.
  • Filter Deals – Filters deals based on the user query.
  • Get many companies – Fetches company records from HubSpot.
  • Filter Companies – Matches companies against the query.
  • Get Contacts – Retrieves contact data from HubSpot.
  • Filter Contacts – Filters contacts using name-based matching.
  • Merge – Combines filtered deals, companies, and contacts into one dataset.

• Generate and send AI response

  • AI Agent – Uses AI to format and structure the CRM data into a readable response.
    • Google Gemini Chat Model – Provides the language model for the AI agent.
  • Send a message – Sends the final response back to the Slack channel.

Why use this?

• Enables instant CRM access directly from Slack without logging into HubSpot
• Simplifies data lookup using natural language queries
• Combines multiple CRM objects into a single intelligent response
• Improves team productivity with faster decision-making
• Easily customizable for additional fields, filters, or AI formatting