Data Scientist / ML Engineer

5 courses · ~12 hours· 2 certifications available

phase 1 — Understand the engine [optional]

Build an autograd engine and neural network from scratch to understand what frameworks hide.

[ ] Understanding Micrograd

You've used PyTorch. This course shows you what it does under the hood — backpropagation, gradient descent, and neural networks built from scalars.

5 sections · 13 lessons

[ ] The Big Picture (2 lessons)
[ ] The Value Class (3 lessons)
[ ] Building a Neural Network (3 lessons)
[ ] Training (3 lessons)
[ ] The Full Picture (2 lessons)

[→ start course]

[ ] Understanding Makemore

Character-level language models from bigrams to WaveNet — the training diagnostics, batch normalization, and manual backpropagation skills you use daily.

5 sections · 15 lessons

[ ] Bigrams (4 lessons)
[ ] The MLP Language Model (3 lessons)
[ ] Activations and Batch Normalization (3 lessons)
[ ] Becoming a Backprop Ninja (3 lessons)
[ ] Building a WaveNet (2 lessons)

[→ start course]

phase 2 — Context as a system

Learn the systems engineering layer for context windows and memory.

[ ] Context Engineering

You already understand tokens and embeddings. This course teaches the systems engineering layer — managing context windows, sub-agents, and memory strategies at scale.

8 sections · 25 lessons

[ ] Tokens and Inference (2 lessons)
[ ] The Real Size of Your Context Window (4 lessons)
[ ] Anatomy of the Messages Array (5 lessons)
[ ] Dynamic Allocation: Tool Calling (2 lessons)
[ ] The Ralph Wiggum Loop (3 lessons)
[ ] Sub-Agents: Managed Runtimes for AI (3 lessons)
[ ] Message Passing: The Erlang OTP of AI (3 lessons)
[ ] Context Management Strategies (3 lessons)

certification: Certified Context Engineer

[→ start course]

phase 3 — Tool calling in depth

Learn agent loop design, parallel execution, and production guardrails.

[ ] Understanding Tool Calling

The advanced patterns and security sections cover agent loop design, parallel tool execution, and production guardrails — the engineering side of your ML systems.

4 sections · 10 lessons