03 Automate: Put repetitive work on autopilot

03 Automate: Put repetitive work on autopilot

03 Automate: Put repetitive work on autopilot

[Embedded HTML content]

Welcome video · 1:25

Welcome video transcript

In the second course, you let AI take action in your apps\. Let's say you handed it one real task: the weekly client report\. It read the new rows in your sheet, wrote the summary, and sent it to the team\. It worked, so you asked again the next week\. And the week after\. In this course, the job starts itself\. Something happens, the workflow runs across your apps, and the result lands, with nobody typing\.

Here's the map again: from chatting with AI to building systems that scale\. Chat is done\. So is Delegate\. Stage three, Automate, is where you put repetitive work on autopilot\. That's this course\.

By the end, you'll describe a job once, and AI builds the workflow\. You'll read what it built, test it, and publish it yourself\. It runs when something happens, or on a schedule, across your apps\. Every run is written down\. And nothing goes live until you say so\.

After that, stage four, Engineer, is where you build AI systems that work reliably at scale\. But that's later\. Right now, you are here\. Stage three\. Let's start\.

About this course

In the second course, you let AI take actions in your apps\. This course takes those one\-off actions and hardens them into recurring workflows\. You build it\. Your workflow runs every time something happens or on a schedule\. In Zapier, a workflow is called a Zap, and you will hear both words in this course\.

Who this course is for

Builders and beginners who finished 02 Delegate and keep asking their AI for the same task\. New to AI in your apps? Start at 02 Delegate\.

What you'll learn

• Why recurring work belongs in a workflow, not another chat\.
• How to describe a job so the agent builds the right Zap\.
• How to read the draft, test it, publish it and change it\.
• How to balance between AI and deterministic steps to cut costs, increase speed, and improve reliability\.

You describe the job in your AI chat and watch it build the Zap in the editor, then test, change and publish it yourself\. The last section follows a real build from ZapConnect 2026 and shows where one AI step is enough, where it gets expensive, and how plain steps and AI steps share the work\.

Lessons

Welcome to Automate

• Your map 1:25

Section 1: Why a workflow

• The case for a workflow 1:55

Section 2: Build it

• Describe the job, the agent builds it 3:32
• Test it, then change it 3:17

Section 3: In practice

• Watch a real build: one AI step runs the whole job 3:34
• Why one AI step on every task is expensive, and where to add determinism 4:18
• When a step breaks, AI steps back in 2:13

Finish

• Check your understanding 4 questions
• Course complete certificate

Start the course

Zapier Academy courses are free\. You sign in with your Zapier account \(a free one is fine\) so your progress is saved and your certificate has your name on it\. Next on the path: 04 Engineer: Build AI systems that work reliably at scale\.