You now have the building blocks: workflows, data and integrations, language models, retrieval, and agents. This module is about where they pay off. The strongest use cases share a pattern: a task that happens often, follows a recognisable shape, and currently eats time because a person has to read something, decide what it is, and move it somewhere else.
Common examples are customer support, where an agent answers routine questions from a knowledge base and passes hard cases to a person; sales and lead handling, where incoming enquiries are scored and sent to the right place; and HR and operations, where documents, messages and approvals move through a set of steps. Notice how often the same parts appear: a trigger, a model that classifies or drafts, a lookup, and a final action.
As you watch, ask of each example what the trigger is, what the model decides, and where a human should still review the result. That habit is how you will pick your own first project. Watch the lessons in order, then continue on to Reliability & Guardrails.
AI Foundations walks through eight practical agents built in n8n. Use it as a tour of what is possible, and for each one note the trigger, the decision the model makes, and the action at the end.
Neil Stephenson builds a support agent that can handle almost any incoming enquiry, using a fictional boxing gym as the example, and explains how to give the agent memory so it can follow a conversation.
freeCodeCamp's course includes complete business examples: a sign-up form that records people in Google Sheets, a WhatsApp alert system, and a CV approval and interview scheduler with a human approval step. Watch the HR and notification examples for this module.
AI Foundations' beginner-to-pro walkthrough of building agents in n8n. Follow along and try to adapt what you build to a task from your own work or study.