Skip to main content

🔄 Make.com Integration with Keymate

Keymate's native integration with Make.com enables powerful no-code automation workflows by connecting your Keymate memory to thousands of other tools. This guide walks you through how to upload files, scrape web pages, and run AI-powered queries — all automated within Make.

🚀 Getting Started​

  1. Log in to Make.com or create a free account.
  2. Use this link to install the Keymate app into your Make workspace.
  3. Authenticate with your Keymate account by creating a new connection inside any Keymate module.

🔧 Available Modules​

Keymate currently offers the following actions:

  • Upload File – Upload a PDF to your Keymate memory
  • URL Scrape – Scrape webpage content and add to memory
  • Query Chunk – Retrieve relevant chunks as plain text
  • Query Chunk (Advanced) – Retrieve chunks with structured metadata
  • Make an API Call – For custom authorized requests

📁 Upload PDFs Automatically (Dropbox Example)​

Goal: Upload new files from Dropbox into a specific Keymate collection.

Scenario Setup:​

  1. Trigger: Dropbox → Watch Files in Folder
  2. Action: Dropbox → Download File
  3. Action: Keymate → Upload File
    • Set the collection
    • Map file data from Dropbox
  4. (Optional): Move file to an archive folder

🧠 Use cases: Client uploads, SOP intake, onboarding kits, etc.

🌐 Scrape Web Content (Google Sheets Example)​

Goal: Scrape and save web content into your Keymate memory using URLs stored in a spreadsheet.

Scenario Setup:​

  1. Trigger: Google Sheets → Watch New Rows
  2. Action: Keymate → URL Scrape
    • Select target collection
    • Map URL column from the sheet
  3. (Optional): Update the row with a "Done" status

💡 Great for: Competitor research, newsletter archives, internal blog scraping

🔍 Semantic Search Query (Query Chunk)​

Goal: Run a query over your Keymate memory and return relevant document chunks.

Scenario Setup:​

  1. Trigger: Custom input (Gmail, Google Form, webhook, etc.)
  2. Action: Keymate → Query Chunk
    • Map query content from the previous step
    • Set target collection (or leave blank to search all)
  3. Action: Use response chunks in next steps
    • Format into a report
    • Pass to LLM (OpenAI/Gemini) for reasoning
    • Email result

🧠 AI Agent Automation (Advanced Use Case)​

Create agent-style workflows where a user message triggers:

  • A structured GPT query
  • Retrieval from Keymate memory
  • AI synthesis using GPT, Claude, or Gemini
  • Action output (email reply, CRM update, Slack message)

Key Modules:​

  • Keymate → Query Chunk (Advanced)
  • OpenAI → Chat Completion
  • Gmail/Slack/Notion/etc. → Output

📌 Tips & Best Practices​

  • PDF Only: Use clean, text-based PDFs for best results
  • Metadata: Use consistent naming and folder structures in Keymate
  • No Deletion Sync: Removing a file from Dropbox doesn't remove it from Keymate — delete manually if needed
  • Organize by Collection: Maintain separate collections for clients, departments, or workflows

🔐 Security Note​

  • All Keymate integrations use secure, authenticated API calls
  • Files are encrypted and processed privately
  • Data is not shared across collections without explicit linking

📚 Resources​