Most people learn n8n backwards: they copy workflows they don’t understand, then the second they need to modify something, they get stuck and quit. In this video I lay out the exact 7-part roadmap I’d follow to go from beginner to building real AI automation systems that actually work in production—no hype. If you put in consistent reps, you can realistically get through this in 1–3 months. We start with foundations: how data actually flows node-to-node (and why you usually don’t need a Loop node), plus the core nodes I use constantly—Set, If, Filter, Summarize, Split Out, Merge, and (yes) Code, but without becoming over-reliant on it. Then we move into triggers (Manual, Chat, Form) and hands-on projects like a task priority sorter, expense categorizer, and contact list cleaner—because messy data is one of the biggest failure points in real-world automations. From there I show you how to connect your first APIs (Google Suite, OpenAI/Anthropic/Gemini) and why the HTTP Request node is non-negotiable. Finally we get into AI fundamentals (prompts, structured JSON outputs, token/context limits), useful AI nodes (extractors, classifiers, summarizers, sentiment), MCP + ChatHub for deploying workflows beyond n8n, and advanced builds like RAG, multimodal workflows, and content repurposing systems.