Pattern is
Everything
Have you noticed the same shapes keep appearing — in data, in price charts? The bell curve in market returns. Gradient descent rolling downhill. A head-and-shoulders pattern at a reversal. This is a map of those recurring structures — 257 interactive, visual references across machine learning, statistics, and markets.
Start here
New here? Start with these
257
Core Topics
257
Core Visualizations
4
Universes
10
Collections
Universes
Pick your universe
Four worlds, ten collections, 257 interactive topics — machine learning, statistics, markets, and the patterns that connect them.
Machine Learning
118 topics across four collections — ML Math, LLM Engineering, MLOps, and Timeseries Engineering. From linear algebra to production deployment, with interactive visualizations and code.
Statistics & Data Analytics
39 topics in The Toolkit — hypothesis testing, statistical foundations, EDA, cohorts, funnels, model evaluation, backtesting, and decision-making. Every topic with interactive visuals, formulas, and Python you can use today.
Market Patterns
100 topics across four collections — Chart Patterns, Technical Indicators, Market Psychology, and Risk & Portfolio. Visual guides to the patterns that move markets.
Pattern Essays
12 short, calm reflections on patterns that appear across the world — the bell curve, the long tail, feedback loops, survivorship bias, fractals, Simpson's paradox, and the Kalman filter. Each with an interactive visualization.
Real-Data Cases
4 practical mini-cases that connect the visual references to datasets, pipelines, metrics, failure modes, and notebook-ready starting points.
Live interactive preview
Every topic is a living diagram
Not screenshots — every reference on this site is drawn live on canvas, with parameters you can drag.
Normal distribution — drag the sliders inside any topic · Try it →
Interactive canvas diagrams
Core formulas explained
JupyterLite notebooks
Real-data mini-cases
Draggable parameters
Dark mode
Zero dependencies
Learning paths
Guided progression
Six curated routes from beginner to advanced — open one to see its steps. Progress is saved locally, no account needed.
Beginner
ML Fundamentals
Start from zero — distributions, metrics, and your first model evaluation. No prior ML experience needed.
~45 min · 5 topics
Build
First Real-Data Model
Move from visual intuition into a practical workflow: dataset, split, model, metric, and failure mode.
~60 min · case + 4 topics
Intermediate
Time Series Mastery
From stationarity to forecasting — the full pipeline for time-dependent data.
~60 min · 5 topics
Advanced
LLM Engineering
From tokenization to RLHF — understand and build with large language models.
~90 min · 5 topics
Beginner
Statistics Foundations
From p-values to Simpson's paradox — the statistical bedrock every honest analysis stands on.
~50 min · 5 topics
Beginner
Market Intuition
Understand how markets move — risk, psychology, and the patterns traders watch. Evidence levels clearly marked.
~45 min · 5 topics
◆ The Pattern
A bell curve in model outputs, a head-and-shoulders on a chart — the same structural principle, different data. Every topic here makes that connection visible.
Beyond the library
Explore differently
A quick tour of the site