Become an AI Engineer — a complete path from LLM basics to production.
Open, build-with-LLMs notes covering the whole stack: how models work, calling them, prompting, retrieval (RAG), agents, evaluation, and shipping to production. No ML-training theory — just what you need to build real AI systems. Free for anyone to study.
🗺️ The AI Engineer Learning Path
Nine stages, beginner to advanced — each builds on the last. Start at the top and work down, or jump to what you need.
0Python for AI
The minimum Python you actually need — APIs, JSON, environments, Git, and a fast Streamlit/FastAPI UI.
Coming soon1LLM Basics
How to think about an LLM, tokens, sampling, temperature & top-p, context windows, and reasoning models.
Read notes →2Calling Models
The API layer — chat completions & roles, streaming, structured outputs, tool calling, and multimodal inputs.
Read notes →3Prompting & LLM Apps
Instruction hierarchy, few-shot, chain-of-thought, prompt templates, context engineering — and shipping an app.
Read notes →4Retrieval (RAG)
The flagship 17-part course — chunking, embeddings, vector search, hybrid retrieval, reranking, and multimodal RAG.
Read notes →5Advanced & Production RAG
Enterprise RAG — hybrid search, reranking, HyDE, CRAG, Self-RAG, Text2SQL, caching, guardrails, and evaluation.
Coming soon6Fine-Tuning & Open Models
When to fine-tune vs RAG, LoRA/QLoRA, Hugging Face Transformers, and running open models like Llama.
Coming soon7Agentic AI
The ReAct loop, tool calling, agent memory, and frameworks — LangGraph, CrewAI, AutoGen — plus multi-agent systems.
Coming soon8AgentOps, MCP & Production
MCP & A2A protocols, tracing & observability, agent evaluation and security, cloud deployment, and monitoring.
Coming soonPrepping for AI Engineer interviews?
100 RAG interview questions with detailed answers — RAG is the #1 most-asked AI-engineering topic in 2026. Start practicing →
Ask My Notes Anything
There's an in-browser RAG bot (bottom-right corner) that answers from these notes and cites its source. Try it!