The curriculum
Foundations, capabilities, and the systems you build from them.
Every capability runs the same engineering loop: dataset, build, eval, trace, fix. Learn the foundation once, then repeat the loop on each capability until it is reflex, and compose the results into production systems.
22 modules are live and free to read, with no account. That is over 12 hours of reading, labs and checkpoints included, and new ones go up as I write them.
Foundation
Cross-cutting rails
Capabilities
- M13ClassificationRead →
- M14RoutingRead →
- M15VerificationRead →
- M16MemoryRead →
- M17PlanningRead →
- M18Tool UseRead →
- M19-1RAG & Retrieval, Part 1Read →
- M19-2RAG & Retrieval, Part 2Read →
- M20Visual Grounding & VLMsRead →
- M21Fine-Tuning
- M22Generation
- M23Extraction
- M24State Understanding
- M30Agents
Compose into systems
- M25AI System Design & Trade-offs
- M26AI Infrastructure & Inference
- M31Orchestration & Multi-Agent Systems
- M27Capstone · Build One End-to-End
Enterprise
- M28Enterprise, Compliance & Governance
The rail through all of it
- M29Working as a Production AI Engineer
Not sure where to start?
M1 is 8 minutes and everything else leans on it. The greyed titles are the modules still being written, and the index stays here as they fill in.
Coming soon
New modules go live as I write them. Get each one in your inbox the day it ships. No spam, just the next lesson.