AI Roadmaps
← All roadmaps
BeginnerRoadmap 05 · 14 steps · 10h 07m of video

AI Engineering

Build real products on top of LLMs: prompting, APIs, structured output, RAG, evals, and local models.

Before this: Basic Python. No ML background needed.

  1. 2h 11m
    Step 1

    How to actually use LLMs

    Practical habits from someone who builds them.

    How I use LLMs

    Andrej Karpathy

  2. 25 min
    Step 2

    Prompting 101

    How Anthropic's own team writes and debugs prompts.

    Prompting 101 | Code w/ Claude

    Anthropic

    dair-ai/Prompt-Engineering-Guide★ 80.2k
  3. 21 min
    Step 3

    Call an LLM from code

    Wrap a model in your own API with FastAPI.

    How To Build an API with Python (LLM Integration, FastAPI, Ollama & More)

    Tech With Tim

    anthropics/claude-cookbooks★ 54.9k
  4. 17 min
    Step 4

    Structured output

    Get reliable JSON out of a model, and why it works.

    Structured Output from LLMs: Grammars, Regex, and State Machines

    Efficient NLP

  5. 4 min
    Step 5

    Embeddings and vector databases

    Turn text into numbers you can search by meaning.

    Vector Databases simply explained! (Embeddings & Indexes)

    AssemblyAI

  6. 7 min
    Step 6

    What is RAG?

    Give a model your data without retraining it.

    What is Retrieval-Augmented Generation (RAG)?

    IBM Technology

  7. 2h 33m
    Step 7

    RAG from scratch

    Indexing, retrieval, and generation, built step by step.

    Learn RAG From Scratch – Python AI Tutorial from a LangChain Engineer

    freeCodeCamp.org

    langchain-ai/rag-from-scratch★ 10.1k
  8. 59 min
    Step 8

    Hybrid search and reranking

    The upgrades that make RAG work on real documents.

    The Complete Guide to Hybrid Search in RAG (BM25 + Embeddings + Reranker)

    Dave Ebbelaar

    NirDiamant/RAG_Techniques★ 30.4k
  9. 9 min
    Step 9

    RAG vs. fine-tuning

    Pick the right tool before you build.

    RAG vs. Fine Tuning

    IBM Technology

  10. 10 min
    Step 10

    Context engineering

    What goes into the context window matters more than the prompt.

    Context engineering explained: What every AI developer should know

    Google Cloud Tech

  11. 28 min
    Step 11

    Evals: common mistakes

    Most AI products fail here. Learn to measure quality.

    LLM Evals: Common Mistakes

    Hamel Husain

  12. 52 min
    Step 12

    Evals on a real product

    Error analysis and LLM judges, worked end to end.

    AI Evaluations Clearly Explained in 50 Minutes (Real Example) | Hamel Husain

    Peter Yang

  13. 14 min
    Step 13

    Run models locally

    Free, private, offline LLMs on your laptop.

    Learn Ollama in 15 Minutes - Run LLM Models Locally for FREE

    Tech With Tim

    ollama/ollama★ 187.7k
  14. 1h 15m
    Step 14

    The AI engineering big picture

    How the job differs from ML engineering, from the author of AI Engineering.

    AI Engineering with Chip Huyen

    The Pragmatic Engineer

    chiphuyen/aie-book★ 18.4k

Build it

Watching isn't learning. Pick one and put it on GitHub.

Where to go next