FASTTRACK
Fasttrack — Complete AI/LLM Crash Course
A visual, self-contained super tutorial: every core concept from your first LLM call to multi-agent orchestration and evals, in 10 lessons.
3-4 hours•10 lessons · 4 phases
What you'll understand by the end
- How to call any LLM from Python with one universal interface
- How to swap between OpenAI, Anthropic, Gemini, and local models without changing code
- How to get typed Python objects instead of raw strings from an LLM
- Why LLMs have no memory and how to simulate it with message history
- How to let the LLM call your Python functions via tools
- How to build a RAG pipeline that answers questions from your own documents
- What an agent is and how it automates the tool-calling loop
- How to give agents memory with checkpointers and manage context windows
- Four agent orchestration patterns: single, sequential, parallel, router
- How to measure and improve prompt quality with evals
Prerequisites
Python 3.10+ installedOPENAI_API_KEY in your .env fileJupyter Notebook or VS Code with Jupyter extension
The journey
Start with the basics, unlock capabilities, build agents, then learn to measure and improve.
Foundation
01-04
Capabilities
05-06
Agents
07-09
Quality
10
All 10 lessons
Foundation
Your first LLM call, provider abstraction, structured output, and memory.
Capabilities
Let the LLM call Python functions and answer questions from your documents.
Agents
Build agents, give them memory, and orchestrate multi-agent systems.
Quality
Measure and improve with evals and prompt engineering.