The AI Learning Navigator: Free Courses Roadmap

Topics covered

• •DAIR.AI's Prompt Engineering Guide — a few days, no coding required, teaches the techniques behind good prompts
• •Microsoft's Generative AI for Beginners — 21 lessons, ~4-6 weeks, explains how LLMs actually generate text, and is one of the few courses on this list with full Russian and Ukrainian translations

Topics covered

• •Hugging Face's Agents Course — free, ~40 hours, walks through smolagents, LlamaIndex, and LangGraph with a graded quiz at the end
• •LangChain Academy — a focused, free course specifically on LangGraph orchestration, best once you already understand basic agent concepts

Topics covered

• •Level 1 — building your first basic workflow: triggers, nodes, and simple data transformations, roughly 1-2 weeks part time
• •Level 2 — complex flows and error handling: sub-workflows, loops, retries, and recovering from failures in production, roughly 2 weeks
• •Advanced AI — building AI agents and LLM chains directly inside n8n, combining deterministic automation with agentic decision-making, roughly 2-3 weeks

Topics covered

• •Marketer or business owner who needs usable output → ComfyUI's docs and tutorials, nothing else required
• •Designer or creative who wants more creative control → ComfyUI first, then Hugging Face's Diffusion Models Course to understand what the settings actually do
• •ML engineer or researcher who wants full mastery → Hugging Face's Diffusion Models Course → fast.ai's 'Stable Diffusion from Scratch'