Skip to main content

The curriculum

73 free, illustrated courses in 10 tracks: mathematics, programming, statistics, machine learning, deep learning, reinforcement learning, applied AI, ML engineering, responsible AI and capstones. One connected path from school maths to research-level AI.

73 courses in 10 tracks hold 671 modules. At four lessons a module that is about 2,684 lessons, or roughly 1,120 hours of study. Every lesson declares what it depends on, so the whole thing is one connected path, not a reading list.

What is live today: 40 modules and 160 lessons. The rest ships in four rings, described in the roadmap.

How the tracks connect

Read left to right. Each track supplies prerequisites to the ones after it.

M · Mathematics

12 courses · L0 to L4

→

P · Programming and Data

8 courses · L0 to L3

→

S · Statistics and Data Analysis

8 courses · L1 to L4

→

ML · Machine Learning

7 courses · L1 to L4

→

DL · Deep Learning

10 courses · L2 to L4

→

RL · Reinforcement Learning

7 courses · L2 to L4

→

A · Applied AI

9 courses · L2 to L4

→

E · ML Engineering

5 courses · L2 to L4

Two tracks run alongside the others from early on:

R · Responsible AI

4 courses · L1 to L4

C · Capstones and Research Practice

3 courses · L1 to L4

The ten tracks

Track Courses Modules Levels What you gain
M · Mathematics Courses 12 Modules 116 Levels L0 to L4 What you gain The maths under every later track
P · Programming and Data Courses 8 Modules 70 Levels L0 to L3 What you gain Python, SQL, data handling and engineering habits
S · Statistics and Data Analysis Courses 8 Modules 75 Levels L1 to L4 What you gain Inference, regression, Bayesian methods, causality, time series
ML · Machine Learning Courses 7 Modules 72 Levels L1 to L4 What you gain Classical ML from the first model to learning theory
DL · Deep Learning Courses 10 Modules 94 Levels L2 to L4 What you gain Neural networks through generative models and theory
RL · Reinforcement Learning Courses 7 Modules 63 Levels L2 to L4 What you gain Bandits through RL for language models
A · Applied AI Courses 9 Modules 81 Levels L2 to L4 What you gain NLP, LLMs, agents, vision, speech, recommenders and classical AI
E · ML Engineering Courses 5 Modules 42 Levels L2 to L4 What you gain Designing, deploying, monitoring and scaling ML systems
R · Responsible AI Courses 4 Modules 34 Levels L1 to L4 What you gain Fairness, interpretability, privacy, safety
C · Capstones and Research Practice Courses 3 Modules 24 Levels L1 to L4 What you gain Projects, paper reproduction, research writing

The five levels

Every course carries a level, so you know what you must already know and what you will be able to do at the end.

Level Name You start with You finish able to Rough equivalent
L0 Name Foundation You start with School arithmetic You finish able to Reads math notation and writes simple Python Rough equivalent Pre-university bridge
L1 Name Core You start with L0 You finish able to Solves standard problems by hand and in code Rough equivalent First-year university
L2 Name Intermediate You start with L1 math and Python You finish able to Derives and implements standard algorithms Rough equivalent Upper undergraduate
L3 Name Advanced You start with L2 in the same track You finish able to Builds and debugs full systems and reads papers Rough equivalent Graduate course
L4 Name Expert You start with L3 in the same track You finish able to Reproduces papers, proves results, extends methods Rough equivalent Research seminar

Every course

M · Mathematics

From school algebra to the stochastic calculus behind diffusion models.

P · Programming and Data

From a first line of Python to GPU computing and data pipelines.

S · Statistics and Data Analysis

How to learn from data with honest uncertainty, from a first histogram to causal inference.

ML · Machine Learning

From a first classifier to generalization theory.

DL · Deep Learning

Build a neural network by hand, then climb to transformers, diffusion models and open theory questions.

RL · Reinforcement Learning

From a slot-machine problem to the RL methods now used to train language models.

A · Applied AI

The ML, DL and RL tracks applied to language, vision, audio, recommendation, reasoning and real business domains.

E · ML Engineering

Turning models into running systems.

R · Responsible AI

The duty of care: fairness, interpretability, privacy, security and safety.

C · Capstones and Research Practice

Where learners prove their skill with projects, reproductions and original work.

How addresses work

Every item has a stable address, so lessons, figures and prerequisites can point at each other.

Item Pattern Example
Course Track code and two digits M02
Module Course and module number M02.4
Lesson Module and lesson number M02.4.3
Figure Lesson and figure number M02.4.3-F2
Interactive Module and interactive number M02.4-X1
Back to top