Learn AI From Scratch: A Complete 38-Lesson Roadmap
A free, structured path to learn AI from absolute scratch — 38 lessons across 6 levels, from ML and neural networks to ChatGPT, RAG, agents, deployment and AI engineering, with diagrams and hands-on code. Start here.
Want to learn AI but don't know where to start? This is your map. We've built a complete, beginner-friendly path — 38 free lessons that take you from "what is AI?" all the way to building, deploying and engineering real AI systems, and even planning a career. No prior AI knowledge needed. Each lesson is written in plain language, with real analogies, diagrams, and hands-on code you can actually run. Bookmark this page and work through it at your own pace.
How to use this roadmap
- Follow the order. Each level builds on the last — start at Lesson 1 and go down.
- Go at your pace. One lesson a day is a comfortable, powerful habit; there's no rush.
- Do, don't just read. Run the code snippets and try the ideas — that's where real learning happens.
- Prerequisites? Just curiosity. A little Python helps later on, and we cover what you need as we go.
Level 1 — Foundations
Understand what AI really is and set up the basics you'll build on.
- Lesson 1: What is Artificial Intelligence? — A simple, jargon-free introduction for absolute beginners.
- Lesson 2: AI vs ML vs Deep Learning vs Data Science — How these buzzwords actually relate.
- Lesson 3: How to Start Learning AI — A step-by-step roadmap for beginners.
- Lesson 4: Python for AI & Machine Learning — The one language you need, and why.
- Lesson 5: Math for Machine Learning — Only what you actually need, explained gently.
- Lesson 6: Statistics & Probability Basics — The intuition behind the numbers.
- Lesson 7: Understanding Data — Datasets, Features & Labels — How machines "see" data.
Level 2 — Core Machine Learning
Learn how machines actually learn from data, and build your first model.
- Lesson 8: Supervised, Unsupervised & Reinforcement Learning — The three ways machines learn.
- Lesson 9: Linear & Logistic Regression — Predicting numbers and categories, with examples.
- Lesson 10: Classification Algorithms (KNN, Trees, Random Forest, SVM) — The everyday ML toolkit.
- Lesson 11: Accuracy, Precision, Recall & the Confusion Matrix — Is your model actually good?
- Lesson 12: Overfitting, Underfitting & Cross-Validation — Why models fail, and how to fix them.
- Lesson 13: Data Preprocessing — Cleaning, Scaling & Encoding — The unglamorous 80% of real ML.
- Lesson 14: Your First End-to-End ML Project — Put it all together on a real dataset.
Level 3 — Deep Learning
Go inside neural networks — the technology behind modern AI.
- Lesson 15: What is a Neural Network? — A simple visual guide to how they work.
- Lesson 16: How Neural Networks Learn — Backpropagation & gradient descent, made intuitive.
- Lesson 17: Convolutional Neural Networks (CNNs) — How AI sees images.
- Lesson 18: Recurrent Neural Networks (RNNs) & LSTMs — How AI handles sequences and text.
- Lesson 19: Transformers & Attention — The architecture behind ChatGPT.
Level 4 — Generative AI
Understand and build with the AI everyone's talking about.
- Lesson 20: Large Language Models — How ChatGPT Works — Tokens, prediction and RLHF.
- Lesson 21: Prompt Engineering — Get far better answers from any AI.
- Lesson 22: RAG — Give AI Your Own Data — Accurate answers with fewer hallucinations.
- Lesson 23: AI Agents — When AI takes actions on its own.
- Lesson 24: How AI Generates Images — Diffusion models behind DALL·E & Midjourney.
Level 5 — Applied & Responsible AI
Build responsibly, and turn your skills into a career.
- Lesson 25: AI Ethics & Bias — A practical guide to responsible AI.
- Lesson 26: AI Project Ideas for Your Portfolio — 9 projects to build and get hired.
- Lesson 27: Careers in AI — Roles, skills and how to break in.
- Lesson 28: The Future of AI — AGI, opportunities, and how to keep learning.
Level 6 — Advanced AI Engineering
Go from using AI to building, deploying and engineering real production systems.
- Lesson 29: Fine-Tuning LLMs — When and how to customize a model (with LoRA).
- Lesson 30: Deploying AI Models (MLOps) — From notebook to production.
- Lesson 31: Build a RAG App Step by Step — A hands-on "chat with your docs" project.
- Lesson 32: How to Evaluate LLMs — Measure if your AI is actually good.
- Lesson 33: LLM Security — Prompt injection and how to defend.
- Lesson 34: Multimodal AI — Models that see, hear and speak.
- Lesson 35: Multi-Agent Systems — AI teams that work together.
- Lesson 36: Cost & Latency Optimization — Making AI faster and cheaper.
- Lesson 37: On-Device AI — Running models locally and privately.
- Lesson 38: Becoming an AI Engineer — Your recap and what's next.
Ready? The hardest part of learning anything is starting. You've already done that by opening this page — now take the first real step below.
Frequently Asked Questions
How do I start learning AI from scratch?
Start with the foundations: what AI is, how it differs from machine learning and deep learning, and a little Python. This roadmap orders 38 free lessons from absolute basics through machine learning, neural networks, generative AI and advanced AI engineering, so you can simply begin at Lesson 1 and work down in order.
Do I need coding or math to learn AI?
Not to begin. The early lessons need only curiosity. A little Python helps once you reach hands-on machine learning, and only light, intuitive math is required — this path teaches exactly what you need along the way, without heavy prerequisites.
How long does it take to learn AI with this roadmap?
It depends on your pace. Doing one lesson a day, you can work through all 38 lessons in about six weeks while genuinely understanding each topic. There is no deadline — the roadmap is designed for self-paced learning, and you can revisit lessons anytime.
Is this AI course really free?
Yes. All 38 lessons are free to read, with explanations, diagrams and runnable code examples. You can work through the entire beginner-to-advanced path, including the advanced AI engineering track, without paying anything.
