Lesson 69 lessons
Connecting AI (ChatGPT) to Make Workflows
The OpenAI module in Make
Add the OpenAI (ChatGPT, DALL-E, Whisper) app module, authenticate with your API key, choose "Create a Chat Completion", and write a prompt referencing data from earlier modules via the field mapper.
Getting structured, parseable output
Instruct the AI to "respond only with valid JSON in this exact format: {category, priority, summary}" so a downstream Text Parser or router can reliably act on specific fields from the response.
A practical example: support ticket triage
New support ticket → OpenAI module categorizes urgency and topic → Router sends urgent tickets to Slack and logs everything to a spreadsheet — turning manual triage into an automated first pass.
Key Takeaways
- The OpenAI module lets you call ChatGPT directly inside a Make scenario.
- Request structured JSON output for reliable downstream parsing.
- AI categorization plus a router builds automated triage pipelines.
- Reference upstream data in your AI prompt using the field mapper.
Build an AI-categorizing scenario
Build a scenario that takes a piece of text (a form response or email), uses the OpenAI module to categorize it, and routes it based on the category.