Exercise 3: Understanding AI Automation

What Is Automation with AI?

Imagine you receive 50 emails every day. You open each one, read it, decide whether it is important, and summarize the key points. That easily takes an hour a day. What if an AI could do that for you?

That is exactly what AI automation is: Recurring tasks that previously required human intelligence are taken over by AI systems. The difference from traditional automation (e.g., an email filter) is that AI can also handle tasks that require understanding and judgment -- like summarizing an email or prioritizing messages.

An Example Workflow: Email Summarization

Let us look at a concrete automation workflow:

Step 1: Trigger

A new email arrives in your inbox.

Step 2: AI Processing

The AI reads the email and creates a summary in 2-3 sentences. It also assesses the urgency (high/medium/low).

Step 3: Action

The summary is automatically written to your note-taking app (e.g., Notion, Apple Notes, Obsidian). Urgent emails receive a flag.

The result: At the end of the day, you have a clear list of all emails with summaries and priorities -- without having read each email individually.

The Most Important Automation Tools

n8n (Open Source)

n8n is an open-source platform for workflow automation. "Open source" means that the code is public and you can even run the tool on your own computer. n8n offers a visual editor where you can click together workflows like a flowchart -- including AI steps.

Zapier

Zapier is the most well-known commercial automation service. It connects thousands of apps to each other and recently started offering AI steps as well. Easier to use than n8n, but paid.

Make (formerly Integromat)

Make is an alternative to Zapier with a visual workflow builder. Especially popular for more complex automations.

Step by Step: Designing a Workflow

Practice designing a simple workflow yourself now. You do not need to build it -- just think it through.

Task: Think of a workflow for the following scenario:

You collect interesting articles from the internet. Every evening, you want to receive a summary of all collected articles as a list.

Think in three steps:

1. Trigger: What triggers the workflow?
2. AI step: What should the AI do?
3. Action: What happens with the result?

A possible solution:

1. Trigger: You save a link in a bookmarking app (e.g., Pocket, Raindrop)
2. AI step: The AI reads the article and creates a summary in 3 sentences
3. Action: In the evening, an email with all the day's summaries is sent out

From Chatbot to Agent: The Next Level

So far, you know AI mainly as a chatbot: You ask a question, the AI answers. But AI can do more than just answer -- it can also act.

An AI agent is a system that:

• Understands a goal ("Summarize my emails")
• Independently plans steps ("First read the emails, then summarize, then save")
• Uses tools (email access, writing to a note app)
• Checks the result and improves it if needed

Agents are the next major development in AI. You have already encountered them in this beginner path -- for example, the Agent Mode in GitHub Copilot or the Background Agents in Cursor. In the advanced path, you will learn to create and configure such agents yourself.

Reflect

In this exercise, you understood the basic principle of AI automation: trigger, AI processing, action. This pattern is the foundation for countless applications -- from email summarization to complex business processes. You do not need to build workflows yourself right away, but understanding the principle is the first step.