The problem
AI education is commonly split between high-level demos and isolated theory, leaving engineers without a coherent path from fundamentals to working systems.
The approach
Organise learning as a build loop: derive the idea, implement it, test it, and keep the resulting artifact. The curriculum spans fundamentals through agent systems and deployment.
My contribution
- Developed and refined the Codeology learning interface and its editorial visual system.
- Improved CV analysis provider handling, failure explanations, and model configuration.
- Contributed across the curriculum application while preserving the build-from-first-principles teaching model.
What exists now
- A public curriculum containing 503 lessons across 20 phases.
- A multi-language learning system covering Python, TypeScript, Rust, and Julia.