Tracks
Pick your track
First chapter of every track is free. Ask the AI anything as you study.
Causal Inference
From "X drives Y" to knowing whether it actually does
Potential outcomes, selection bias, DAGs, and the full quasi-experimental toolkit — matching, diff-in-diff, regression discontinuity, instrumental variables, synthetic control — in plain English, grounded in the canonical texts. Ends with the AI-era skillset: ML meets causal inference, and auditing causal claims with AI as your skeptic.
Data Analysis
From raw table to defensible insight — the full analysis workflow
The end-to-end craft of analyzing data: understanding a dataset’s grain, data quality and missing values, transformations and joins, distributions and outliers, cohorts and funnels, and charts that don’t lie — finishing with directing AI through a full analysis and auditing the result.
Experimentation
Design, size, run, and read A/B tests you can defend
Grounded in Oehlert’s design-of-experiments text and Bruce & Bruce’s Practical Statistics: why randomization works, permutation-test intuition, power and MDE, peeking and multiple comparisons, variance reduction, factorials and interactions, and bandits.
Machine Learning
Hands-on machine learning in Python — from preprocessing to deep learning
Build, evaluate, and judge real models in Python, grounded in Raschka’s Python Machine Learning, Goodfellow’s Deep Learning, and Mitchell’s and Alpaydin’s ML texts: data preprocessing, regression, classification, clustering, association rule learning, reinforcement learning, NLP, deep learning, dimensionality reduction, and model selection.
Python Coding
Write real Python in the browser and pass real tests
Read pandas transformations, spot silent data-loss bugs, then flip from reader to operator: describe the analysis you need, let AI write the notebook, and check its work like a reviewer who can’t be fooled.
SQL Coding
Write real queries in the browser and pass real tests
Understand SQL deeply enough to catch a query lying to you — then use AI to write your own queries and know, not hope, that they’re right. By the end you pull your own numbers instead of waiting on the data team.
Statistics
From sample vs. population through regression — worked by hand and in Python
Statistics from the ground up, grounded in Statistics Unplugged, Practical Statistics for Data Scientists, and Statistics for Business and Economics: descriptive statistics, sampling and probability, discrete and continuous distributions, estimation and confidence intervals, hypothesis testing, and regression — plus the aggregation traps that fool dashboards.