AI & ML Quick Path

AI and machine learning, from zero to industry projects

A complete, step-by-step guide for people new to AI. It explains every concept in plain words, shows how companies use it, and gives working examples in both Python and .NET (C#). Where a lesson uses maths or statistics, a short collapsible panel shows the main formula with simple worked examples. Everything you need is on these pages; there are no outside links to follow.

50 lessonsAbout 67 hours of study7 parts3 visual roadmaps8 capstone projects11 real company cases

Start lesson 1See the 12-week plan

Seven parts that build on each other, from the big picture to production 1. Big picture2. Data3. Machine learning4. Deep learning5. Language models6. AI applications7. Production
Each part uses the ones below it. You can go back to any part at any time.

The seven parts

  • Part 13 lessons, about 4 h

    The big picture

    What AI and machine learning are, how a real AI project runs from idea to production, and how to set up Python and .NET on your own laptop.

    0 of 3 completed

  • Part 24 lessons, about 6 h

    Data

    Every model is only as good as its data. Learn what data looks like, how companies collect and label it, how to clean and explore it, and how to turn it into features.

    0 of 4 completed

  • Part 39 lessons, about 12 h

    Machine learning

    How models learn from examples, the main types of problems (predict a number, pick a category, find groups, spot anomalies, forecast, recommend), the main algorithms, and how to measure and improve results.

    0 of 9 completed

  • Part 49 lessons, about 12 h

    Deep learning

    Neural networks and how to train them, then the model families used for images, text, audio, generation, decision-making and connected data.

    0 of 9 completed

  • Part 58 lessons, about 11 h

    Large language models

    How large language models work, how to design one, how to train a small one on your laptop and scale it to big hardware, how to fine-tune and run models locally, and how to evaluate them.

    0 of 8 completed

  • Part 612 lessons, about 16 h

    Building AI applications

    Build real products on top of models: calling LLMs, prompting, embeddings, RAG, agents and tools, multimodal apps, evaluation, security and responsible AI.

    0 of 12 completed

  • Part 75 lessons, about 8 h

    Production

    Run AI and ML reliably for real users: serving models as APIs, MLOps, monitoring, cost and scale, and reference architectures that combine .NET and Python.

    0 of 5 completed

How every lesson is built

All 50 lessons use the same eight sections in the same order, so you always know where to look.

SectionWhat you get
In plain wordsThe idea explained in plain words; every new word is defined.
Real-world exampleWhere a real business uses it and what it decides.
How it worksThe steps in order, with what goes in and what comes out.
Python exampleRunnable code, with the real output it printed.
.NET (C#) exampleThe same idea in C#, with the packages it needs.
Applied Maths - StatisticsCollapsible panels with the main formula, a plain-words meaning and two worked examples for each maths idea the lesson uses (where it applies).
Maths behind itThe names of the maths topics involved, so you can recognise them in documentation and job descriptions.
Common mistakesWhat goes wrong in real projects.
Quick checkQuestions with hidden answers to confirm you understood.

Choose how to start

Get a complete grip (recommended)

Follow Parts 1 to 7 in order. Read "In plain words" and "How it works" first, run the Python or .NET example in your main language, then answer the quick check before moving on.

  1. What AI, machine learning, deep learning and generative AI are
  2. How an AI or ML project runs from idea to production
  3. Data basics: types, features, labels and splits
  4. How a model learns: loss, training and overfitting
  5. Classification: picking a category
  6. Neural networks in plain words
  7. How large language models work
  8. RAG: answering from your own documents
  9. Tools and agents with human approval
  10. Serving a model as an API

About 12 weeks at 8–10 hours a week, plus capstones (study estimate).

I want to build AI apps first

Start with how models learn and how LLMs work, then go straight to Part 6. Come back to Parts 3 and 4 when you need to train your own models.

  1. What AI, machine learning, deep learning and generative AI are
  2. How a model learns: loss, training and overfitting
  3. How large language models work
  4. Calling an LLM from your application
  5. Prompt engineering that works in production
  6. Embeddings and vector search
  7. RAG: answering from your own documents
  8. Tools and agents with human approval
  9. Evaluating AI applications
  10. Security for AI applications

About 4–6 weeks at 8–10 hours a week to a first working AI app (study estimate).

I want to train models first

Work through data and machine learning, then deep learning, then train a tiny language model on your laptop. Add Part 7 to put your models into production.

  1. How an AI or ML project runs from idea to production
  2. Data basics: types, features, labels and splits
  3. Cleaning and exploring data
  4. Feature engineering: turning raw data into model inputs
  5. Regression: predicting a number
  6. Classification: picking a category
  7. Measuring models honestly
  8. Neural networks in plain words
  9. Train a small language model on your laptop
  10. Serving a model as an API

About 6–8 weeks at 8–10 hours a week to a trained, served model (study estimate).

What you will be able to build

  • Capstone 1 Machine learning

    Customer churn prediction service

    A subscription business wants a daily list of customers most likely to cancel, plus a live score on the customer-care screen, so the retention team can call the right people first.

  • Capstone 2 Machine learning

    Store demand forecasting

    A retail chain wants next-14-day demand per product per store so it orders enough stock without filling the warehouse.

  • Capstone 3 Machine learning

    Visual defect inspection

    A factory wants to flag scratched or dented parts from camera images so inspectors only check suspicious items.

  • Capstone 4 AI application

    Company knowledge assistant (RAG)

    Employees ask questions about HR and IT policies; the assistant answers only from approved documents, cites the passage and says "I don’t know" when the documents do not cover it.

  • Capstone 5 AI application

    Order-returns agent with human approval

    A shop wants an assistant that can look up orders, check the returns policy and prepare refunds, but a staff member must approve every refund.

  • Capstone 6 Language models

    Train a small language model on your laptop

    Understand language models by building one: train a small transformer on a text collection on your own laptop, then plan how the same code would scale on GPUs.

  • Capstone 7 Language models

    Private assistant running fully on a laptop

    A law office wants a document assistant that never sends text outside the building: a local model summarises and answers questions over local files.

  • Capstone 8 Language models

    Fine-tune a small model with LoRA

    A support team wants a small model that answers in the team’s own short, polite style and returns the same JSON fields every time. You fine-tune a small open model with LoRA on a few hundred examples and prove with a test set that it beats the plain prompted model.

How lessons are labelled

  • Beginner needs only the lessons listed as prerequisites.
  • Intermediate builds on several earlier lessons.
  • Advanced needs Parts 1 to 3 and some practice.
  • Expert combines many parts into full systems.

Honest about the code. Every lesson states whether its code was run. Python outputs shown in lessons were copied from real runs on 2026-11-02. The C# examples were written carefully but could not be compiled in the build environment, and are labelled that way. Durations are study estimates; quick checks are self-checks, not certificates. Progress is saved only in this browser (see progress).