English companion article · Published

Start with the system, not the terminology

This course follows the material from my Agentic AI workshop, recorded again for people who want to learn at their own pace. It begins with AI history and familiar AI tools before moving into prompting and the basics of large language models. That order matters: using a chatbot gives you a starting point, but building an agent requires understanding how information and actions move through a system.

Context, loops and useful memory

The next layer is context engineering: deciding what information the model needs for the current task. Loop engineering brings in repeated action and feedback. The course then compares AI workflows with agents before introducing vector databases, short-term context and longer-term memory.

Retrieval-augmented generation, or RAG, connects the discussion to relevant information stored outside the immediate conversation. MCP and APIs introduce the connections through which an agent can work with other tools. These concepts are presented as foundations, so you can understand their roles before going deeper into implementation.

Turn the concepts into a small workflow

The practical n8n section begins at 59:37. Watch it after the conceptual chapters, then choose one bounded task you understand well. Identify its input, the information it needs, the tools it can use and the result you want to inspect. That gives you a much clearer learning exercise than trying to automate an entire business at once.

Use the chapter links below to revisit individual concepts as you build. The goal is to connect the vocabulary to a working process you can explain and improve.

← Explore all articles & videos