How this platform teaches

Retrieval practice, spacing, interleaving, immediate feedback — old ideas from learning research, applied seriously to a curriculum this deep.

The methods here are not new: the things learning research has backed for decades. What’s unusual is applying them without compromise to a curriculum with the standards of a university programme and the practical bent of actual engineering work.

First principles, actually01

Nothing is assumed. Variables, derivatives, pointers, distributions, instruments — all taught from zero, and no formula shows up before the reason it exists.

Prerequisites are enforced, not suggested02

Lessons unlock through a prerequisite graph covering 283 lessons. If a lesson says you need something, the thing you need is on the platform — you shouldn’t have to leave mid-lesson to go find it.

Mastery over exposure03

Score 80%+ on the knowledge check and finish the exercises, and the lesson is “passed”. To call it “mastered” you have to hold onto it — the material comes back later, and you have to land it again.

Spaced retrieval04

Every lesson generates flashcards, scheduled with SM-2 (the Anki algorithm). Quiz questions you miss join the same queue. Reviewing a card after it was due is what upgrades a lesson from passed to mastered.

No fake anything05

No employment promises, no “learn C++ in 21 days”, no metrics invented to look impressive. If a module isn’t finished, the platform says so. Every number you see was measured.

How a lesson works

Every lesson follows the same 20-part structure so you always know where you are

  1. 1Title & estimated time
  2. 2Prerequisites (enforced)
  3. 3Learning objectives
  4. 4A real motivating problem
  5. 5Intuitive explanation
  6. 6Formal explanation
  7. 7Visual/interactive demonstration
  8. 8Step-by-step derivation
  9. 9Worked examples with verified numbers
  10. 10Guided exercise
  11. 11Independent exercise
  12. 12Coding implementation where applicable
  13. 13Common mistakes
  14. 14Real-world application
  15. 15Quantitative finance application
  16. 16Knowledge check (server-scored)
  17. 17Summary
  18. 18Flashcards created
  19. 19Further practice links
  20. 20References

Assessment types

Nine ways you prove it — none of them guessable

Multiple-choiceMultiple-selectNumeric problemsShort-answer reasoningCode-output predictionFull judged programming problemsTimed assessmentsScreening simulationsCumulative retrieval checks

The bar is deliberately strict: you can’t guess your way through, and time-on-page counts for nothing. That can feel harsh. It beats the alternative.

Content standards

Every lesson is validated by an automated schema gate (283 lessons, 65 modules): quiz answers verified present, math delimiters balanced, code samples syntactically plausible, interactive components registered, prerequisites resolvable. Mathematical examples carry hand-verified numbers; finance examples use clearly-hypothetical data. References prefer primary sources — cppreference, OpenStax, Python and NumPy documentation, PostgreSQL and Linux docs — and never substitute for the lesson itself.