v1.0.0 — the course at launch

The complete first release of linearly: linear algebra for people building AI, free for everyone.

What ships

• 17 lectures (00 through 15, including 7.5), each with a full plain-language article, a slide deck, computed figures, an interactive laboratory, and the same code in NumPy, PyTorch, JAX, and TensorFlow
• 127 figures under one color canon: green column space, orange row space, cyan null space, purple left null space, blue for the action of the matrix — every coordinate computed, every Excalidraw source committed
• 20 interactive components, from a finite state machine you can drive to an eigenvector explorer
• The roadmap (a staged path with a study method and honest hours) and the library (a hard-curated shelf with insider notes, every link verified)
• Responsive from a 360px phone up, with a 40px touch floor on every control
• A contributors page rebuilt from this repository's history on every deploy

Licensing: free to learn from, closed to commercial use. Code under PolyForm Noncommercial 1.0.0, content under CC BY-NC-SA 4.0.

Contributing: the course is built to be grown. CONTRIBUTING.md carries the writing canon, the drawing rules, and the full technical workflow; AGENTS.md is the rulebook for AI agents.

In memory of Aqshin.