top of page

How to Build Your First AI Agent With Make.com

Person in a cap, looking up, with text "Build Your First AI Agent with Make.com" and AI-related icons on a blue-white background.

Meet Your New AI Teammate: No Coding, No Stress

Imagine if you could build an AI agent to handle repetitive tasks — like responding to customer queries or managing classroom Q&A — without writing a single line of code. Welcome to your future-self’s favourite life hack, made possible with Make.com, Tally, and Google Docs.


Whether you’re a startup founder, a time-strapped teacher, or a multitasking content creator, this guide walks you through a powerful automation that works for you while you sip your coffee (or run another meeting).


Ready to supercharge your workflow with some friendly automation? Let’s dig in.


🛠️ Tools You’ll Need

  • Make.com – The automation powerhouse

  • Tally – For your form submissions

  • Google Docs – As your info database

  • An OpenAI account – Your AI brain


🚀 What You’re Building

You’ll create a no-code automation where:

  1. A user submits a Tally form.

  2. Make.com reads the form data.

  3. An AI agent (powered by OpenAI) reads instructions from a Google Doc.

  4. It composes a personalised response based on both.

  5. The AI sends that response back to the user — via email.


It’s like building your own digital concierge. Or…an incredibly polite robot assistant.


🧠 Step-by-Step: Building Your AI Agent


🧩 Part 1: Setting Up the AI Agent

  1. Head to Make.com and log in.

  2. Click AI Agents on the left menu.

  3. Add a new connection and select OpenAI (or another LLM provider if you’re feeling experimental).

  4. Choose a name, select a model (GPT-4 is usually a good bet), and enter your prompt. Example:

    “Answer user questions clearly, kindly, and based only on the information in the document.”


💡 Pro tip: Keep your prompt focused but friendly. You want your agent to act like a helpful colleague, not a corporate automaton.


📄 Part 2: Feeding the AI Some Brains (aka Google Docs)

  1. In Make, go to the Scenarios page.

  2. Add the module: Get content of a document.

  3. Connect your Google account and choose the document your agent will reference.

  4. Send this document data to the next module (this passes info to the AI).

  5. Add “Scenario Outputs” and map them correctly — you’re teaching your agent where to look.


📬 Part 3: Crafting the Response Email

  1. Create a new scenario.

  2. Add input modules for:

    • The user’s email address

    • Email subject

    • Email body (this will be the AI’s response)

  3. Add the email-sending module.

  4. Connect your email account (Gmail works well).

  5. Map the email fields to match your previous inputs.


🎨 Optional flourish: Customise the subject lines to reflect the question’s topic for a nice UX touch.


📝 Part 4: Connect It All With Tally

  1. Add the Tally module: Watch New Responses.

  2. Choose your form and map the questions you’re collecting (e.g. user email, question, etc.).

  3. Plug in the AI Agent module next — this is where the magic happens.

  4. Then connect the email scenario you created.


🧠 Don’t forget: You can provide additional explanations to the AI to shape how it responds.


✅ Final Steps

  • Set the scenario to "On Demand" or schedule it as needed.

  • Test it. (Yes, break it a little — better now than when it’s live.)

  • Hit Save. Take a moment to admire your digital handiwork.


Who’s This For?


This setup is brilliant for:

  • Customer support: Auto-respond to FAQs with doc-backed answers.

  • Educators: Personalised student support without the inbox chaos.

  • Project managers: Collect form inputs, cross-check with a project doc, and keep your team in the loop.


Basically, if you're tired of repetitive queries — this automation will save you hours every week.



🎓 Real Talk From Netropolitan Academy

At Netropolitan Academy, we believe digital tools should be empowering, not overwhelming. Whether you’re a seasoned dev or an AI-curious creative, automation like this gives you more time to do what you love — think, teach, build, create.


You don’t need to be a coder to be a creator in this space.


So give it a try. Build something. Tinker. Learn. Share.


Let your robot assistant handle the admin while you focus on the magic.


💬 We’d Love to Hear From You

Have you tried building your own AI agent? Got stuck on a step? Want to show off your workflow? Drop your thoughts in the comments or connect with us at Netropolitan.xyz.


👋 About the Author


Işıl Benek


Computer Engineering Student | Tech Enthusiast | Digital Explorer


Interned with Ecodation, Europe Coding School, and Velo Games. Proficient in Python, Java, HTML/CSS. Proud data analyst, creative coder, and community builder.


Follow me on LinkedIn to connect and swap ideas!

Comentarios

bottom of page