AI Engineering Course: an 18-module curriculum kept as markdown in a repository

A curriculum in markdown, with one blog post per lesson

The AI Engineering Course is a free and complete course to learn AI Engineering step by step, from Machine Learning, Neural Networks and Transformers through to LLMs, Fine-Tuning, RAG, AI Agents, LLM Inference, Evaluation, AI Safety and AI System Design. Every lesson comes with a detailed blog, and many lessons come with a video.

The About section gives the scale: it is a free, structured and step-by-step curriculum with 18 modules and 146 or more in-depth lessons, and every lesson is a detailed blog that explains one concept in simple words with examples, diagrams and math wherever it is needed.

The repository layout is as small as you would expect for that. There is a README.md, a LICENSE, an assets/ directory for the images, and a.gitattributes file. The primary language recorded for the repository is Markdown, which is the whole point. There is no src/ directory, no tests, no package.json and no build step, because the deliverable is the writing.

The project is Apache-2.0 licensed, the default branch is main, and the last push was on 2026-10-22. The repository publishes no GitHub releases, which for a course is the right shape, since there is nothing to release that is not already in the tree.

There is also a note that the course will continue to grow as the author writes more blogs and creates more videos on new topics, so the module list in the README is a snapshot rather than a promise of completeness.

The module list is the curriculum, and it is sequenced deliberately

The table of contents in the README lists eighteen modules, and the order is the argument the course makes.

Module 0 is Must Know, and Module 1 is Machine Learning Foundations, followed by Module 2 on Deep Learning and Neural Networks. Module 3 is Generative AI and the Transformer Architecture, Module 4 is How LLMs Generate Text, and Module 5 is Modern LLM Architecture. Module 6 covers Types of Language Models, and Module 7 is Training, Fine-Tuning and Alignment.

Then the engineering half. Module 8 is Prompt Engineering and Context Engineering, Module 9 is Vector Search and Retrieval-Augmented Generation, and Modules 10 and 11 are AI Agents and Agentic Systems and Agentic Engineering and Agent Frameworks. Module 12 is LLM Inference Engineering, Module 13 is Evaluation and Observability, and Module 14 is AI Safety and Security. Module 15 covers Multimodal AI and Generative Models, and Module 16 is AI Infrastructure, Deployment and System Design. Module 17 is Frontier Ideas in AI, and Module 18 is Prepare for AI Engineering Interviews.

The last two are the interesting ones for judging the intent. A module on frontier ideas keeps the course honest about what is unsettled, and a module on interviews tells you the course is aimed at people who need to be hired for this work.

There is a glossary at the end, the AI Engineering Key Concepts Glossary, and an FAQ section. The glossary matters more than it sounds, because a course that introduces terms in order needs a lookup for the ones you half remember.

The About section also frames the whole thing in one line, saying this is the course the author wishes he had when he started learning AI Engineering.

What the author means by AI engineering

The What is AI Engineering section defines the discipline as building real-world applications and systems on top of AI models, especially Large Language Models, and then describes the role in a way that is worth reading closely because it draws a line around the job.

An AI Engineer does not always train a model from scratch. Instead, an AI Engineer knows how these models work from the inside, picks the right model, gives it the right context, connects it to tools and data, makes it fast and cheap to run, measures if it is doing a good job, keeps it safe, and ships it to real users.

The summary is three clauses: AI Engineering equals understanding the model, plus building on top of the model, plus running the model in production. And the note says this is exactly what the course teaches.

The reading of the module list follows from that definition. Modules 3 to 7 are the understanding part, and they are the ones a working practitioner would most likely already know. Modules 8 to 14 are the building and running part, and those are the ones that decide whether the application works. Module 16 is the shipping part.

The intended audience list is unusually long and specific, which is itself informative about the market. It includes AI Engineer, Gen AI Engineer, LLM Engineer, Agentic AI Engineer, AI Agent Engineer, Forward Deployed Engineer, AI Solutions Architect, AI Platform Engineer, Applied AI Engineer, Machine Learning Engineer, MLOps Engineer and LLMOps Engineer. The Who is This AI Engineering Course For section adds software engineers moving into AI, backend, mobile and frontend developers building AI-powered products, machine learning engineers and data scientists going deeper into LLMs, RAG and agents, students and freshers, engineering managers and tech leads, and anyone preparing for AI engineer interviews.

Every lesson follows the same four promises

The About section lists what every lesson is, and the four items are more like a style guide than a feature list.

Every lesson is free to read, with no sign-up and no paywall. Every lesson is written for beginners, with no jargon and no assumptions. Every lesson is detailed, going from why do we need it, to how does it work step by step, to where is it used in the real world. And every lesson is in order, each one building on top of the previous one.

That last point is the one that makes a course different from a pile of articles. The ordering is a claim about dependency: you cannot understand why a particular technique works before you understand the problem it solves, and the curriculum is the argument for that ordering. The same three-part structure inside each lesson, why then how then where it is used, is a claim that an idea is only learned when it is placed in a context.

Written for beginners and detailed are in some tension, and the resolution the author gives is the inclusion of math wherever it is needed alongside examples and diagrams. So a lesson can start with no assumptions and still include the equations once the intuition is there.

The format also means the content is durable in a way a video series is not. A lesson is a blog post, so you can search it, link to a section, and read it at the pace you want, and the videos are an addition rather than the only form.

It is also a funnel, and the README says so

The I teach at Outcome School section is the part to read if you are deciding how much time to spend here, and it is short.

The course is completely free to read. If you want more practical and in-depth learning with the author in live classes, there is a paid live program at Outcome School, an AI and Machine Learning live class linked from the same page.

The author is named and positioned in the README as the Founder of Outcome School, with links to a YouTube channel, an X account, a LinkedIn profile and the author's GitHub account, and the same set of links for the school. The homepage field points at the outcomeschool.com program page rather than at a repository-hosted course, which matches the structure: the free material is the funnel and the program is the paid half of the same curriculum.

That is not a criticism so much as a fact about how to read the course. The free course is a complete ordered curriculum in the sense that it covers the whole arc, and it is not a substitute for the live classes in the sense that it cannot watch your code fail. Both statements are in the README.

The audience list and the interview module both point the same way, at a career transition. If that is your situation, the course is unusually well suited to the goal, because the ordering is the thing a self-directed learner most needs and least often gets.

Editorial conclusion

Use the AI Engineering Course if you want one ordered, free written path through AI engineering that goes from the transformer internals to serving and evaluation, and you are prepared to read prose rather than run code. Do not expect a project to clone and build, because the repository is a course in markdown with no exercises or sample code. Before you start, check the module list against what you already know, since the sequencing assumes a beginner and the early modules will be familiar to a working engineer, and note that the author runs a paid live program on the same material, which is worth knowing when you weigh the free text against it.

Frequently asked questions

Official sources

1. amitshekhariitbhu/ai-engineering-course on GitHub
2. Issues
3. License: Apache-2.0
4. Project website
5. README