Building AI Agents

About this course

A chatbot answers your question.

An agent, on the other hand, can use tools, make decisions, and keep going until a job is done.

This course teaches you how to build those agents, starting with what they actually are and ending with agents that can research, organizee, monitor, draft, coordinate and work together.

You will learn the thinking behind a good agent, then build them in Claude, n8n and Claude Code using real tools, real information and clear guardrails.

If you have taken Introduction to Building with AI, great. It is a useful foundation, but absolutely not a prerequisite. You can also start here with no prior coding experience.

The goal is here is to learn how to give AI the right job, the right context, the right tools and the right amount of freedom.

What you'll do

Understand what separates an AI agent from a chatbot, and how the agent loop actually works

Learn the four parts of an agent: the model, tools, memory and context

Use the AGENT checklist to turn a vague idea into a clear, safe brief

Build productivity agents for coordination, creativity, research and rehearsal

Build a visual agent in n8n and connect it to tools such as Google Calendar and Slack

Build a Reddit listening agent in Claude Code that finds useful conversations, filters them and drafts replies for you to review

Add memory, permissions, budgets, testing and guardrails so your agents behave predictably

Turn one specialized agent into a broader Growth Agent, then build a Chief of Staff that can coordinate work across agents

What you'll leave with

A practical framework for deciding what should and should not become an agent

Several agents you have built and tested yourself

A working Reddit listening system that can find relevant conversations, log them and prepare draft replies for your review

A reusable Growth Agent with skills you can continue adding over time

A simple multi-agent structure where one agent can coordinate work across others

A much clearer understanding of tools, memory, context, autonomy, permissions and guardrails

Most importantly, you will leave with the ability to look at a repetitive or messy piece of work and ask: could I build an agent for this?

Who is this for

People who have moved past asking AI questions and want to start giving it jobs.

Your starting point might be that you're constantly juggling assignments, trying to do the work of a small team, or simply someone curious about what AI can actually do beyond a chat box.

You do not need to be technical or know how to code. You do need to be willing to experiment, inspect what the agent produces, and improve it when it gets something wrong.

If you have a repetitive task, a messy workflow or an idea for something you wish AI could handle for you, you have enough to start.

Building AI Agents

About this course

What you'll do

What you'll leave with

Who is this for