Data science roadmap: from Python to generative AI

A free, step-by-step data science roadmap. Work through five stages in order, from your first line of Python to machine learning, deep learning and generative AI. Each stage builds on the one before it, and each has its own route map that shows only the chapters you need.

1. Stage 1: Python

   The language everything else is written in, from the basics to concurrency.

   20 free chapters · about 10 hours · Fundamentals, Objects and data types, Operators, Flow control, Built-in data types, Functions…

2. Stage 2: Python libraries

   The data toolkit: arrays, tables and charts.

   Coming soon · NumPy arrays, NumPy maths and broadcasting, pandas Series and DataFrames, Cleaning data with pandas, Grouping, joining, reshaping, Matplotlib…

3. Stage 3: Machine learning

   Every classic algorithm built from scratch, then checked against scikit-learn.

   26 free chapters · about 17 hours · Maths foundations, Data and features, Model evaluation, Linear and logistic regression, Trees and ensembles, SVMs and k-NN…

4. Stage 4: Deep learning

   Neural networks, from one neuron to transformers.

   Coming soon · Neurons and activations, Backpropagation by hand, Training and optimisers, PyTorch basics, CNNs for images, RNNs and LSTMs…

5. Stage 5: Generative AI

   LLMs, prompting, RAG and LangChain.

   Coming soon · How LLMs work, Tokens and embeddings, Prompt engineering, Vector databases, RAG, LangChain basics…