I have the privilege of teaching first-semester students at the university.
In practical and applied classes on Artificial Intelligence and Machine Learning, we use n8n as a central tool. Six years ago, I began working with automation using Integromat (now Make.com) and APIs, but the landscape has changed dramatically with the arrival of language models (LLMs) and recent advances in AI. Today, tools like n8n seem key to teaching these concepts.
Unlike Make.com, which runs in the cloud, I set up my own n8n server using Docker on Linux, which allows me full control over the entire process: from installation to production.
Among the activities I teach is the creation of AI Agents with memory and tools, integrating APIs like Google Gemini. These agents, for example, can check the current time with GetCurrentDate, access Google Calendar, or connect to other relevant services.
We also explored how to put a chatbot built with n8n into production and customize its interface to resemble WhatsApp.
As part of the hands-on practice and assignments, we developed workflows ranging from a customer service chatbot for a retailer to website scraping to automatically generate content and publish it on social networks like X (formerly Twitter).
