Curriculum
A path, not a playlist.
Four tracks, twenty-three weeks end to end if you follow it linearly. Each module ends with a lab that is graded automatically, so you find out whether the concept stuck before moving on.
Track 01 — Foundations
4 weeks- 01Python for data work
- 02Linear algebra you'll actually use
- 03Probability & statistics
- 04Notebooks, environments, and reproducibility
Track 02 — Core modeling
6 weeks- 01Regression and regularization
- 02Trees, forests, and boosting
- 03Neural networks from scratch
- 04Honest evaluation and error analysis
Track 03 — LLM engineering
8 weeks- 01Transformer architecture
- 02Fine-tuning & adapters
- 03Retrieval & RAG systems
- 04Agents & tool use
- 05Evaluation & deployment
Track 04 — Production
5 weeks- 01Serving and autoscaling
- 02Monitoring and drift
- 03Cost control
- 04Safety, governance, and documentation
The LLM track in detail
Our most popular track. Every module pairs a short concept lesson with labs you run against real models.
- 01Transformer architecture6 labs
- 02Fine-tuning & adapters8 labs
- 03Retrieval & RAG systems7 labs
- 04Agents & tool use9 labs
- 05Evaluation & deployment5 labs