I’ve absolutely loved creating 6 Udemy courses on Gen AI and Agents. It’s brought me much joy to share my enthusiasm for the world of LLMs. I never expected this, but 400,000 people are taking the courses from 190 countries. It’s bewildering and I’m super grateful.

I get this question a lot: in what order should the courses be taken? Here’s my definitive answer with the curriculum for becoming an AI expert.

Which course to start with

• Start with AI Builder if you want to use products to create AI Agents
• Start with AI Coder if you want to use AI Agents to create products!

The complete curriculum with links

Sidenote:
I updated some of these course names recently for clarity:
AI Engineer Core Track was previously “Master LLM Engineering”
AI Engineer Agentic Track was previously “Agentic AI”
AI Engineer Production Track was previously “AI in Production” and “MLOps”

Which order to take the courses?

All the courses are designed to be complementary and can be taken in any order. You should pick the order that most resonates with you. All things being equal, it might be most natural to follow the exact order in the table above; then each course will build on the last.

I looked at actual job descriptions for AI Engineering roles in designing this curriculum. Employers typically look for expertise from all these tracks, making this a comprehensive roadmap to becoming a proficient AI engineer.

Be listed as a Proficient AI Engineer

I’ve set up a directory for those that complete the full curriculum and wish to be listed. It links to your Digital Twin from the courses, so that prospective clients or employers can learn about you and register their interest. If you’d like to be included, please register here.

Details of the 6 courses

AI Builder: Create Agents, Voice Agents & Automations in n8n

This is perhaps the most satisfying of all my courses! In 3 intensive weeks, you’ll become an expert at making Agents with n8n and ElevenLabs.

Non-technical people will be astonished by what we achieve with these fantastic platforms. And actually, technical people will be astonished, too. An integrated Agentic App that would have taken weeks can now be delivered in hours.

I pack a ton into 3 weeks: APIs, Agents, Context Engineering, Security and of course MCP. And as always: we build several high impact commercial products, with a grand finale that will leave you extremely satisfied.

AI Coder: Vibe Coder to Agentic Engineer in 3 weeks

Most of my courses involve using code to make AI Agents.
But not this one.
This course is the opposite! This course covers using AI agents to make code!

In 3 weeks, I will take you on a rollercoaster. We start gently with Cursor, Copilot, Codex and Antigravity, covering best practices around AGENTS.md and context management. Then the pace picks up as we YOLO and build complete products from scratch.

Then in week 2 it gets real. We move from Vibe Coding to Vibe Engineering at a professional grade. We roll up sleeves and get deep into Claude Code, covering Slash Commands, Checkpoints, MCP, Skills, Plugins, Ralph Loops. We also look at OpenCode and Amp. By the end of the week we can raise an issue in Jira and watch in astonishment as our Coding Agent picks it up, builds it, tests it, and pushes changes to GitHub.

And in week 3 it gets mad! Sub-agents, multi-agents, Claude Agent Teams, GSD, Gas Town – we’ll have 10 agents hammering away while we get coffee, running in Yolo on Sandboxes. We explore what’s going on at the frontier with Claude Agent SDK, Cowork and OpenClaw. We’ll end with a capstone project that will make your jaw drop.

The course is intended for all backgrounds, whether you’re new to programming or a senior architect. There’s something for everyone. The only requirement is to buckle up – you’ll be in for a wild ride!

AI Leader: Generative AI & Agentic AI for Leaders & Founders

This one is not actually a course at all! It’s a briefing, not a course, designed for Business Leaders and Founders, but very relevant for technical and product people too. A business workshop on Gen AI for people at all companies from early stage startups to global enterprises.

I get a steady stream of ‘blah blah blah’ presentations from management consultants in my inbox but they don’t tackle the true thorny problems facing this industry. So I wanted to make a gritty, real-world briefing, with actionable material.

There are 3 parts: be an AI strategist, be an AI decision-maker, be an AI leader. Each part ends with a concrete toolkit that you can put directly into action.

While this is a commercial briefing, I strongly recommend it for tech people too. It’s a super-power for an AI Engineer to demonstrate business thinking and a commercial mindset.

AI Engineer Core Track (refreshed for 2026)

This is an intensive 8 week program to build LLM expertise. It starts with APIs and models, followed by selecting open-source models, RAG, fine-tuning with QLoRA, and building Agent platforms from scratch. It includes seriously meaty commercial projects with a shocking result in week 7.

From Oct 2025 – Jan 2026, I updated all the videos and labs to be a completely updated and refreshed version.

AI Engineer Agentic Track

A 6 week comprehensive journey into the most fascinating area of AI right now: autonomous Agents. It covers OpenAI Agents SDK, CrewAI, LangGraph, AutoGen and MCP. It includes some astounding projects, and some crazy projects, culminating in a Trading Floor of equity trader agents, partly coded by a team of software engineering agents. It’s mind-blowing!

AI Engineer Production Track

In 4 weeks, you’ll be able to deliver full stack AI applications to Vercel, AWS, GCP and Azure; you’ll be comfortable working with Docker containers, Terraform and Github Actions. We go deep on AWS and you’ll be confident with many AWS services, but in particular Lambda, App Runner and of course Bedrock and SageMaker.

You’ll be able to build RAG pipelines, and deploy multi-agent systems including MCP. You’ll have core MLOps expertise. And, you’ll be equipped to deploy SaaS applications with user management and subscriptions. Most importantly, you’ll do it at enterprise-grade, positioned for scalability, security, resiliency, monitoring and observability.

All that, in 4 weeks! It’s an intensive course: action-packed, a bit grueling in places, but always immensely satisfying.

Thank you!

If you have any more questions or ideas for me, please do reach out at ed@edwarddonner.com or LinkedIn (I occasionally subtly mention I love connecting). Or, sign up here:

Finally, let me share my YouTube playlist on LLMs here – along with the Udemy content, I have so much fun making these, and I hope to make more soon..