Interactive Workbenches

The academy is a browser-based practice environment, not just readings. Reusable, data-driven components let you manipulate every core object. All run offline; Python executes for real in a sandboxed Web Worker (numpy/scipy/pandas), lazy-loaded only when you open a coding workbench.

Function & Calculus Explorer

Interactive graphs, secant→tangent derivatives, Riemann integrals, Taylor approximation (Phases 2–3).

Probability Laboratory

Distribution explorers, sampling, LLN and CLT demonstrations (Phase 7).

Linear Algebra Laboratory

2D linear maps, eigenvectors, conditioning, covariance/PCA (Phases 4–5).

Optimization Laboratory

Convexity, gradient descent, Newton, conditioning, portfolio problems (Phases 3, 10).

Brownian Motion Laboratory

Path simulation and quadratic-variation visualization (Phases 8–9).

Monte Carlo Laboratory

Convergence, confidence intervals, variance reduction, option pricing (Phases 13–14).

Time-Series Workbench

Series plots, ACF/PACF, differencing, stationarity (Phase 12).

Derivatives & Volatility Workbench

Black–Scholes pricer, Greeks explorer, implied vol, smiles (Phases 14–15).

Order-Book & Execution Simulator

Limit-order book, spread, slippage, market impact (Phase 16).

Python Coding Workbench

Real Python in a sandboxed offline Pyodide worker with test-driven grading (all phases).

Security model
Learner code runs in an isolated Web Worker with execution timeouts, output caps, no DOM access, and no network. The runtime is bundled locally. See docs/CODING-WORKBENCH-SECURITY.md.