The AI infra curriculum
Follow a sequence of lessons from service foundations to GPU systems and model operations. Each course connects the explanations to related concepts and practice questions, so you can apply what you have learned. Every course has free lessons you can read without an account.
Learn how AI infrastructure works through explanations, worked examples and diagrams, then practise applying it to interview questions.
Not sure which course to choose? Follow the learning journey.
Start at the first unfamiliar step. The route teaches systems decisions and model operations; programming syntax is outside its scope. Follow the core path, then choose a specialty.
- 01 · SERVICE FOUNDATIONS
From Model to Service connects networking, cloud resources, identity, data, evaluation and retrieval to a deployment plan.
- 02 · UNDERSTAND THE WORK
Foundations of AI Infrastructure explains GPUs, request phases, memory, latency and the arithmetic behind capacity.
- 03 · TUNE THE SERVICE
Inference Engineering develops engine selection, cache management, batching, precision and serving-fleet decisions.
- 04 · OPERATE THE PLATFORM
Fleet and Reliability covers scheduling, storage, networking, telemetry, failures and multi-tenant operations.
Training specialty. Training at Scale follows updates, parallel layouts, checkpointing and recovery.
Kernel specialty. Kernels and the Machine is optional for readers pursuing GPU programming and profiling.
Each course has a free first lesson. Lesson labels distinguish free reading, free-account access (sign in) and Premium before you open them. Premium projects turn the explanations into design memos, capacity plans and runbooks.