Make & Zapier: Integrating AI into Your Existing Tools
You connect AI to your daily tools via Make or Zapier without writing code.
Why this lesson matters
Not everyone wants to, or can, self-host N8N. And sometimes you just want to add an AI step to an existing workflow quickly. Zapier and Make are the two most popular no-code automation platforms, and both have good AI integrations.
In this lesson you learn when to pick which platform, and you build two concrete examples: an email summary through Zapier and a document generator through Make.
Zapier: the simplest option
Zapier is the most used automation tool in the world. Its strong points:
- The biggest range of integrations: from Gmail to Salesforce to obscure niche tools
- AI Actions: built-in AI steps, where for some integrations Zapier supplies the model itself
- Extremely easy to get into: if you can fill in a form, you can use Zapier
- Build in plain language: you describe what you want and Zapier sets the step up for you
Zapier has a free plan with 100 tasks/month. For serious use you start at $20/month (750 tasks). AI Actions are available from the free plan up.
Zapier example: email summary to Slack
Scenario: Automatically summarise every important email with Claude and post the result in a Slack channel.
┌─────────┐ ┌──────────┐ ┌──────────┐
│ Gmail │──→│ Claude │──→│ Slack │
│ Trigger │ │ Summarise│ │ Message │
│ (new) │ │ │ │ │
└─────────┘ └──────────┘ └──────────┘Step 1: Trigger: New email in Gmail (filter: only from important senders)
Step 2: Claude AI: Prompt:
Summarise the following email in max 3 bullet points.
State whether any action is needed (yes/no).
If yes, describe the action in 1 sentence.
Email:
{{body_plain}}Step 3: Slack: Post the result in #team-updates with this format:
Keep reading
Leave your name and email address and you can read the rest
You get the whole lesson right away, and every other lesson stays open after that. No password, no confirmation email.
📧 New email from: {{from_name}}
Subject: {{subject}}
Summary:
{{claude_output}}You build this without a single line of code. Do count on needing your own API key: with most integrations you connect your own account, and the model's costs then run through you.
Make: more visual and more powerful
Make (formerly Integromat) is Zapier's visual competitor. Where Zapier is linear (step 1 → step 2 → step 3), Make offers:
- Visual canvas: drag modules and draw connections
- Routers and filters: split workflows based on conditions
- Iterator and aggregator: process lists and combine results
- Lower price at volume: $9/month for 10,000 operations
- Error handling: extensive error handling per module
Make is the better choice when you have workflows with branches, loops or complex logic. Zapier is better for simple "if this, then that" automations.
Make example: form to document to email
Scenario: A customer fills in an intake form → AI generates a personalised proposal → the document is emailed to the customer.
┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐
│ Typeform │──→│ OpenAI │──→│ Google │──→│ Gmail │
│ Form │ │ Proposal │ │ Docs │ │ Send │
│ │ │ generate │ │ Create │ │ │
└──────────┘ └──────────┘ └──────────┘ └──────────┘Step 1: Typeform trigger. The customer fills in name, company, needs.
Step 2: OpenAI module. Prompt:
Write a professional proposal based on
these customer details:
Name: {{name}}
Company: {{company}}
Needs: {{needs}}
Budget: {{budget}}
Structure: introduction, approach, timeline, investment.
Max 1 A4 page, professional tone.Step 3: Google Docs. Create a new document with the AI-generated proposal. Use a template for the layout.
Step 4: Gmail. Send the document as a PDF attachment to the customer with a personal cover email.
Always check AI-generated documents before they go to customers, certainly at the start. Add an approval step: have the document emailed to yourself first, check it, then forward it by hand. Only automate fully once you trust the quality.
Comparison: Zapier vs Make vs N8N
| Criterion | Zapier | Make | N8N |
|---|---|---|---|
| Difficulty | Easy | Medium | Medium to hard |
| Price (entry) | $20/month | $9/month | Free (self-host) |
| Integrations | The most | Many | The fewest |
| AI integrations | Good | Good | Excellent |
| Visual building | Limited | Very good | Good |
| Complex logic | Limited | Good | Very good |
| Self-hosting | No | No | Yes |
| Writing code | No | Limited | Yes (JavaScript) |
| Best for | Quick, simple zaps | Visual workflows | Technical teams |
Which platform when?
Use this decision tree:
Do you have technical knowledge?
├── No → Zapier
├── Yes → Do you want to self-host?
│ ├── Yes → N8N
│ └── No → How many workflows?
│ ├── Few (< 10) → Zapier
│ └── Many (> 10) → Make (cheaper at volume)You don't have to choose: use Zapier for simple automations and N8N or Make for complex AI workflows. Combine platforms through webhooks.
What is the biggest advantage of Zapier over Make and N8N?
When do you choose Make over Zapier?
In the Make example, why is an approval step important for AI-generated documents going to customers?
Want to keep your progress, get the practice files and sign up for the free evening? Create a free account. All you do is click a link in your email.
Create a free account