AllDayCode
Get started

Live For students, universities and engineering teams

Learn AI the way it's actually built.

One path from Python and the maths of ML to neural networks, LLMs and production AI. Read clear notebook lessons, watch the ideas move in interactive visualizations, then solve real problems with real data.

Start learning For institutions
Notebooks you run yourself Instant-feedback quizzes Progress saved as you go
Courses
20
Notebook lessons
476
Interactive visualizations
91
Real-world builds
3

One platform

Three ways to learn, all in one place.

Academy

Notebook courses from Python and linear algebra to LLMs, agents and MLOps — with quizzes in every lesson.

See the courses

Visualizations

Watch algorithms run step by step, change the parameters, and read a short guide to the theory behind each one.

See them move

Build Yourself

Solve one real problem several ways — from simple rules to fine-tuned models and LLMs — and decide what you'd ship.

See a build

Academy

One path from Python to production AI.

Each course builds on the last, so you understand why things work — then learn to ship them.

Read

Every lesson is a carefully written notebook: plain-language explanations, the maths, and real code with its output.

Run

Copy the code or download the notebook and run it in Jupyter, VS Code or Colab — on your own machine.

Check

Short quizzes give instant feedback with explanations, and your progress is saved as you go.

Getting ready

  • 00 Python & Data Toolkit for AI

Foundations

  • 01 Math for ML: Linear Algebra & Calculus
  • 02 Probability & Statistics

Core machine learning

  • 03 Machine Learning
  • 04 Neural Networks & Deep Learning
  • 05 Time Series & Forecasting

Language & LLMs

  • 06 Natural Language Processing
  • 07 Embeddings & Vector Search
  • 08 Large Language Models: How They Work

Building with AI

  • 09 Building LLM Applications
  • 10 AI Agents & MCP
  • 11 LLM Evaluations
  • 12 Production AI: MLOps & LLMOps

Advanced ML & research

  • 13 Advanced Machine Learning: Probabilistic Models, Kernels & Learning Theory
  • 14 Probabilistic Machine Learning
  • 15 Generative Models: VAEs, GANs, Flows & Diffusion
  • 16 Reinforcement Learning & Control

Electives

  • 17 Evolutionary Computation & Genetic Algorithms
  • 18 Responsible AI, Security & Governance
  • 19 AI for Tech Leaders & Product Teams

Choose your path

Recommended routes through the courses for where you are today.

University students

AI Engineer

The full journey: Python, maths, ML, deep learning, NLP and LLMs — with a capstone in every course.

~210 hours

Enterprise developers

LLM Application Developer

Embeddings, how LLMs work, RAG, agents, evals, guardrails and production — ship LLM features safely.

~75 hours

Data & ML practitioners

Data Scientist → ML Engineer

Statistics for ML, classical and deep learning, then MLOps to take models to production.

~85 hours

Leads, PMs & managers

Tech Leader

What AI can and can't do, product economics, running AI projects, risk and governance.

~25 hours

Visualizations

See the ideas move.

91 interactive visualizations, from gradients and eigenvectors to attention, diffusion and reinforcement learning. Each one comes with a guide: how to read it, the theory behind it, and things to try.

  • Animations — play an algorithm step by step, forwards and back.
  • Explorables — drag points, change parameters, watch what happens.
  • Explainers — short read-along stories with a live picture.

Lessons link straight to the visualizations that explain them.

Animation

A neural network bends space

Explainer

Self-attention, step by step

Explorable

Gradient descent on a loss surface

Animation

Diffusion: noising and denoising

Build Yourself

Real problems. Real data. Several ways to solve them.

Each build takes one problem a company really has — routing support tickets, reading invoices, moderating comments — and climbs a ladder of approaches: simple rules, classic ML, embeddings, fine-tuned transformers and LLMs. You run every notebook, compare them on one scoreboard, and decide what you would ship.

  • Public, real-world datasets
  • Accuracy, speed and cost side by side
  • A notebook for every approach
  • Ends with a ship decision

Scoreboard

Route support tickets to the right team

Banking77 · 13,069 real bank-support messages, 77 intents (CC BY 4.0)

  • Keyword rules · Baseline 73.0%
  • Naive Bayes · Classic ML 93.9%
  • TF-IDF + logistic regression · Classic ML 96.7%
  • Sentence embeddings + logistic regression · Pretrained embeddings 97.5%
  • Fine-tuned MiniLM · Fine-tuning 97.5%
  • LLM few-shot (400-ticket sample) · LLM 88.7%

Team macro-F1 on held-out data · the lesson: With labelled data, cheap trained models beat the LLM at a tiny fraction of its cost; a confidence threshold with a human …

For universities, colleges & companies

Bring a complete AI curriculum to your students.

We work with universities, colleges and companies to deliver AI training: ready-made courses, a private space for your cohort, and hands-on workshops.

Talk to us

  • Your own organization

    A private space for your class or team. Invite students by email and give faculty their own roles.

  • Choose the courses

    Decide which courses each cohort sees, including courses made only for your institution.

  • The whole platform

    Your students get the Academy, every visualization and the Build Yourself projects.

  • Workshops & training

    Hands-on sessions delivered together with your faculty or engineering leads.

Start with your first lesson today.

Sign in, pick a course, and your progress is saved from the first page.

Get started