The full curriculum
23 learning units live now. 176 exercises. Every concept taught through a live playground, not slides.
Concepts
74 exercisesMeet the LLM: How Prompting Actually Works
FreeInteract with an LLM for the first time and build a working mental model of how prompting works.
What Is an LLM? Building Your Mental Model
Explore the behavioral architecture of a Large Language Model.
How LLMs Are Built: Training, Fine-Tuning & Model Families
Understand how LLMs are trained and the differences between model families.
Tokens: The Atomic Unit of Every LLM Interaction
Master tokenization, token limits, and how they affect cost and context.
Context Windows: Memory, Limits & What Gets Forgotten
Understand context window limitations and how to work within them.
Temperature & Sampling: Dialing Consistency vs. Creativity
Control randomness and creativity with temperature and sampling parameters.
Hallucinations: Why LLMs Lie Confidently and How to Fight Back
Understand why models hallucinate and strategies to reduce it.
Safety, Ethics & Alignment: Guardrails, Attacks & Responsible Deployment
Learn about AI safety, bias, and responsible deployment practices.
Engineering
57 exercisesThe Anatomy of a Prompt: 5 Building Blocks
FreeDecompose prompts into Task, Context, Examples, Format, and Constraints.
Clarity and Specificity: Killing Vague Prompts
FreeRewrite vague, fuzzy prompts into concrete, specific, measurable instructions.
Showing, Not Telling: Examples and Few-Shot Prompting
FreeUse examples to control output style, format, and quality without writing long instructions.
Shaping the Output: Format, Length, and Tone
Reliably produce output in a target format, length, and tone.
The Prompting Loop: Iteration and Debugging
Treat prompting as an iterative loop—diagnose failures and fix the right layer.
Roles, Personas, and Reasoning Scaffolds
Apply persona prompts and reasoning scaffolds appropriately.
Capstone: Build It End-to-End + Bridge to Module 2
Synthesize every Module 1 concept into one real, multi-part prompt.
Advanced
45 exercisesRAG: Giving LLMs a Knowledge Base They Were Never Trained On
Build Retrieval-Augmented Generation systems to extend LLM knowledge.
From Prompts to Systems — The Advanced LLM Landscape (2026)
Map the full advanced LLM ecosystem across architectural layers.
Advanced Prompt Engineering — Chain-of-Thought, Meta-Prompting & Prompt Chaining
Master advanced prompting techniques for complex tasks.
System Prompts & Structured Outputs — Production-Grade Prompt Architecture
Architect production-ready system prompts with structured JSON outputs.
LLM APIs in Production — OpenAI, Anthropic & Gemini Deep Dive
Build resilient, production-grade LLM API integrations.
Advanced RAG Architecture — HyDE, Re-ranking, Hybrid Search & Multi-hop Retrieval
Build production-grade RAG systems with advanced techniques.
Vector Databases & Embeddings — Choosing, Indexing & Scaling for Production
Design and scale vector database infrastructure.
LLM Agents & Tool Use — ReAct, Function Calling & Multi-step Reasoning
Build autonomous agents with tool use and multi-step reasoning.
See something you want to learn?
The first 50 exercises are free. Open the playground and start learning.
Try Module 1 — FreeStop watching. Start tinkering.