Your First Automation Workflow

So far you have used AI by sitting at the keyboard: you open a chat, type a prompt, copy the result. That is powerful, but it still needs you for every single run. Automation is the next step — you set up the work once, and it runs itself whenever a condition is met. No coding, no servers, just connecting a few tools and letting them talk to each other. This lesson demystifies what a workflow actually is and walks you through building your first one.

What an Automation Workflow Actually Is

Strip away the jargon and an automation workflow is just one sentence: "When this happens, do that." When a form is submitted, draft a reply. When a payment lands, send a thank-you. When a new row is added to a spreadsheet, summarise it. That is it. Every automation you will ever build — simple or complex — is a chain of those "when this, do that" statements.

You do not write code to create them. You use a no-code automation platform that provides the connections between apps and an interface to wire them together. The three most popular are Zapier (easiest to start), Make (more visual and flexible), and n8n (the most powerful, and the focus of a whole separate course on this site). For your first workflow, any of them works.

Triggers and Actions: The Two Halves

Every workflow has exactly two kinds of building block, and understanding them is most of the battle:

• The trigger — the event that starts the workflow. "A new email arrives." "Someone fills in my contact form." "It's 9am on a Monday." A workflow always begins with exactly one trigger.
• The actions — the steps that run after the trigger fires. "Send the text to ChatGPT." "Save the result to a Google Doc." "Notify me on Slack." You can chain as many actions as you like.

The AI part of your workflow is simply one of the actions: a step that sends some text to ChatGPT or Claude with a prompt and passes the answer along to the next step. That is the bridge between everything you learned in lessons 1–3 and a system that runs on its own.

Your First Build, Step by Step

Let's build something genuinely useful: when a new enquiry comes into your contact form, AI drafts a personalised first reply and saves it for you to review. You stay in control — AI does the typing, you do the sending. Here is the shape of it:

Trigger: New contact-form submission (e.g. from a Google Form, Typeform, or your website form).
Action 1: Send the submission details to ChatGPT or Claude with a drafting prompt.
Action 2: Save the AI's draft reply into a Google Doc or email draft labelled "To Review".
Action 3 (optional): Send yourself a notification that a draft is ready.

In your chosen platform you would pick the trigger app, connect your form, then add an "AI" or "OpenAI" action and paste the prompt below. Every platform names the buttons slightly differently, but the logic is always trigger → AI action → save. Build it once and every future enquiry gets a thoughtful first draft within seconds.

The Prompt That Powers the AI Step

The quality of the automation lives entirely in this prompt — everything you learned in Lesson 3 applies. Notice how it assigns a role, gives the task, sets the format, and adds constraints. The [[form fields]] are placeholders the platform fills in automatically from the trigger:

Because the workflow only drafts the reply, you read it, tweak anything that needs a human touch, and send. You get the speed of automation with none of the risk of a robot emailing your clients unsupervised.