Courses

Artificial Intelligence

Machine learning, deep learning, NLP, and real-world AI applications

  • 9 lessons
  • 5.5 hours
  • Intermediate
  • Graded labs: 2
Start course Course map & resources

What you will learn

  • Set up a reproducible notebook environment and run a model end to end for the first time.
  • Prepare numeric data in Python and compute the maths ML relies on: vectors, dot products, derivatives and gradient steps.
  • Train, evaluate and compare a supervised model, and explain overfitting using a train/test split.
  • Describe how a neural network learns (forward pass, loss, backpropagation) and implement a single neuron by hand.
  • Explain how text becomes tokens and embeddings, and why that lets language models search, summarize and chat.
  • Explain how convolutions let a model see images and pick a sensible vision pipeline for a task.
  • Outline the path from notebook to production: versioning, serving, monitoring and data drift.
  • Use generative models responsibly: write effective prompts, check outputs and recognise hallucination and bias.

How you are assessed

You prove AI understanding by predicting what code will do, not by memorising terms: quizzes check the concepts, labs make you implement the maths, and a challenge puts it under pressure.

To pass this track
  • Complete every Foundations lesson
  • Average on first-try quizzes: at least 70%
  • Graded lab exercises to pass: 2
  • Challenges to solve: 1
  • Lessons

    Cover the core toolkit

    Complete the setup, Python, maths, machine-learning and deep-learning lessons.

  • Quizzes

    Reason about models

    Average at least 70% on first-try quizzes on overfitting, loss functions and evaluation metrics.

  • Lab exercises

    Implement the maths

    Pass the normalization and gradient-descent labs: no libraries, just Python.

  • Challenges

    Solve under pressure

    Solve one data or algorithm challenge that needs a model-style calculation.

Curriculum

Module 1 · 5 lessons

Core ML toolkit

Python, maths and your first supervised and deep models.

0/5
  1. 1.Setup & Prerequisites — Your First Notebook

    Before your first model: install a code editor, install Python, get a real AI notebook running (Google Colab or Anaconda), and create free accounts on Kaggle and Hugging Face. Step-by-step — ready for everything else in this roadmap.

    25 min 1 exercise 1 quiz
    Try in Code Lab
  2. 2.Python for AI — Your First Tools

    Python is the language of AI. NumPy for arrays, Pandas for data, Matplotlib for visualization, and Jupyter notebooks for experimentation. You cannot do AI without Python.

    40 min 2 exercises 1 quiz Graded labs: 1
    Lab Try in Code Lab
  3. 3.Math for ML — Less Than You Fear

    Linear algebra (matrices, vectors, eigenvalues), statistics (probability, distributions, Bayes), and calculus (derivatives, gradients). You don't need a PhD — you need intuition for these concepts.

    35 min 2 exercises 1 quiz Graded labs: 1
    LabChallenge Try in Code Lab
  4. 4.Machine Learning — Learning from Examples

    Supervised learning (regression, classification), unsupervised learning (clustering), model evaluation, overfitting, cross-validation, and Scikit-learn. This is the core of AI.

    40 min 1 exercise 1 quiz
    Challenge Try in Code Lab
  5. 5.Deep Learning — Neural Networks

    Neural networks, backpropagation, CNNs, RNNs, and transformers. Implement in PyTorch or TensorFlow. Deep learning powers image recognition, NLP, and generative AI.

    35 min 2 exercises 1 quiz
    Try in Code Lab
Module 2 · 2 lessons

Language and vision

How machines read text and understand images.

0/2
  1. 6.NLP & LLMs — How Machines Read

    Natural Language Processing: tokenization, embeddings, transformers (BERT, GPT), fine-tuning, and working with LLM APIs (OpenAI, Anthropic, Hugging Face). The most in-demand AI skill in 2024.

    40 min 2 exercises 1 quiz
    Try in Code Lab
  2. 7.Computer Vision — Teaching Machines to See

    Image classification, object detection (YOLO), image segmentation, and OpenCV. Computer vision is used in self-driving cars, medical imaging, and security cameras.

    35 min 1 exercise 1 quiz
    Try in Code Lab
Module 3 · 2 lessons

Production and generative AI

Ship and monitor models; use generative AI with judgement.

0/2
  1. 8.MLOps — From Notebook to Production

    Deploying, monitoring, and maintaining ML models in production. MLflow for experiment tracking, Docker for containerization, and cloud ML platforms (SageMaker, Vertex AI).

    25 min 1 exercise 1 quiz
    Try in Code Lab
  2. 9.Generative AI — Machines That Create

    Prompt engineering, RAG (Retrieval Augmented Generation), building AI agents, LangChain, vector databases, and image generation (Stable Diffusion). The bleeding edge of AI in 2024.

    30 min 1 exercise 1 quiz
    Try in Code Lab

Run a real example

This is the kind of code you will write. Run it, then change it and run it again.

Python

Machine learning in miniature: the model nudges two numbers downhill until its errors shrink. Change the learning rate to 1.0 and watch it fail.

Open in Code Lab

Recommended challenges

Put the skills to work against a puzzle with a flag or a test suite.

See all challenges

Badges

Earned automatically from your progress on this device.

  • First Model

    Locked

    Finish your first lesson and run a model.

    0 / 1

  • Maths Comfortable

    Locked

    Complete every core-toolkit lesson.

    0 / 5

  • Gradient by Hand

    Locked

    Pass 2 graded lab exercises.

    0 / 2

  • Overfit Spotter

    Locked

    Score 80% or more on the first try in 4 lesson quizzes.

    0 / 4

  • AI Practitioner

    Locked

    Complete every lesson in the track.

    0 / 9

Certificate of completion

Lessons completed: 0 / 9

Complete every lesson to unlock your certificate.