Learn AI EngineeringThrough Building
Six in-depth tracks, each with a crash course, a full fundamentals track, and a real project — from RAG to raw PyTorch.
Seven Tracks
RAG Systems
Retrieval, reranking, and evaluation — crash course, a 17-page fundamentals track, and a hospital patient assistant project
AI Agents
Tools, the loop, planning, memory, multi-agent systems, and safety — crash course, a 13-page fundamentals track, and a support operations agent project
MCP
Tools, resources, prompts, transports, and OAuth — current with the 2026-07-28 spec — crash course, a 16-page fundamentals track, and a remote server project
MLOps
Serving, containers, orchestration, CI/CD, registries, monitoring, and LLM gateways — crash course, a 16-page fundamentals track, and a production gateway project
Deep Learning
Neural networks, the transformer, LoRA, quantization, and DPO — from raw PyTorch — crash course, a 16-page fundamentals track, and a from-scratch transformer project
Small Language Models
Distillation, quantization, on-device deployment, and speculative decoding — crash course, a 16-page fundamentals track, and an in-browser assistant project
HuggingFace Ecosystem
The Hub, transformers, PEFT, TRL, accelerate, and diffusers — crash course, a 16-page fundamentals track, and a fine-tune-to-Space project
Every Track, Same Shape
Read the whole thing in 30 minutes, go deep on what you need, then build
Crash Course
~30 minutesOne page covering the whole topic at working depth, with links to go deeper
Fundamentals
13–17 pagesA full track, from first principles through production and failure modes
Projects
hours to daysA complete, real build — or a scoped brief to build yourself