Learn AI
A world-class, prerequisite-driven curriculum. Start anywhere — every concept links to what you should know first and what to read next.
The base ideas behind all of AI: models, training, and inference.
How machines learn patterns from data.
Neural networks and the architectures that power modern AI.
Attention and the architecture behind today's language models.
Large language models: tokens, context, prompting, and tool use.
Turning meaning into vectors you can compare mathematically.
Storing and searching embeddings at scale.
Retrieval-Augmented Generation: grounding models in real information.
AI systems that reason, plan, and use tools to reach goals.
The Model Context Protocol connecting models to tools and data.
Designing what goes into the model's context window.
How AI systems remember across turns and sessions.
Models that understand and generate images, audio, and video.
Adapting models to your data and tasks.
Measuring whether an AI system actually works.
Tracing, logging, and monitoring AI systems in production.
Defending AI systems against injection, abuse, and leakage.
Serving models efficiently: latency, throughput, and cost.
Designing complete, production-grade AI systems.