AI Coder: Complete Claude Code & Coding Agents Course (Udemy.com)

It can be improved further, I would recommend subscribe to Ed Donner YouTube channel also, there he has video how to deal with large codebase. I am only giving 4.5 rating because I did not like lecture related GasTown Multi-agent orchestration system. I will never use GasTown because of their own vocabulary. Overall, it's a 5-start rated course.

This is the third course from Ed's series I've finished. It so worth the ride. I starting the Production track, so I already know that I can deploy any project I build to a live web site. As a software developer, this is both frustrating and the secret sauce. Code that was boilerplate and routine blocked most of personal projects, but instead of getting smart on code, I got smart on AI and beat the code to the punch. My biggest concern is the big, conservative companies will hire developers and look at agentic coding as the boogeyman to be met with torches and pitchforks --or brooms in SouthPark. I'm still doing JDK 7 to 8 upgrades instead of microservices migrations. This genie is out of the bottle and isn't going back/

This course completely changed the way I approach coding agents. I came in knowing about tools like Claude Code and Codex, but without really understanding how to use them effectively. I left with a much better understanding of planning, context management, PRs, debugging, agent workflows. The most importantly, I learned how to think and work alongside an LLM. What I particularly appreciated was the distinction between Vibe Coding and Vibe Engineering. Building an impressive MVP is one thing; keeping the code traceable, understandable, editable, testable, and maintainable is another. The course helped me understand both the incredible potential of these tools and their current limitations. The hands-on projects made the difference. I even built my own version of a trading workstation using my portfolio, historical risk/return analysis, efficient frontiers, and rebalancing suggestions. For me, this wasn't just a course about AI coding tools. It was an introduction to a fundamentally different way of engineering software. Highly recommended.

I enjoyed this course so much. I'm genuinely impressed by how agentic AI was able to build commercial-quality projects at such speed. I was watching the agent teams and GSD build out the trading workstation, and it was the highlight for me. The usage limits were the only frustrating part, but I had the patience to work around them and the project still came out amazingly. Being on Windows added a few hurdles the videos don't cover, though nothing that stopped me finishing. Ed teaches very clearly and everything was easy to understand, and the resources page was genuinely useful whenever I got stuck. This course was a great match for me and I can't wait to start his another AI course.

For me this has been one of the easiest AI agentic courses I had the pleasure of going through. Starting from mostly zero knowledge of Agentic AI to completing this course I now feel ready to take on my own personal exciting projects. Ed Donner - the teacher is very engaging, clear and it´s a pleasure and fun following along his instructions. Explains perfectly on how to do things! I will continue onward with more of Eds courses soon.

The instructor was knowledgeable, enthusiastic, and engaging throughout the course. The content was well organized and covered an impressive breadth of important AI coding tools, concepts, and workflows. It was a practical and highly useful course that I can immediately apply to real-world projects. This instructor has already become one of my favorites, and I look forward to taking more of his courses.

Hi Ed, I am a non-tech person. I was following well at week 1 and week 2, but week 3 was tough for me. But I enjoyed the process of learning this course with you. Every time I heard a new tech word, I checked and found out what it is. As more tech words I learned, I started feeling that my knowledge to AI is broaden. I enjoy this learning journey. I will keep gridding.

The course was much longer than it needed to be. Only about 10% of the content felt truly valuable while the rest focused on things that weren’t very relevant. Too much time was spent demonstrating and commenting on the application itself, and not nearly enough on the actual AI concepts and implementation. I’ve been working in the IT industry for a long time, and spending more than 16 hours on this course felt like a waste of time. I also have concerns about the impression this course may create for beginners. Completing this course alone should not be considered sufficient preparation to call oneself an AI engineer or AI professional. If such a course becomes a substitute for real engineering experience and gives candidates an advantage over experienced engineers, it raises serious concerns about the current state of the industry.

Ed Donner does a fantastic and enthusiastic job of presenting on AI Coding concepts with relatable information and hands-on exercises that are easy to follow. I DO recommend using VS Code as it helps follow what the instructor is doing for 90%+ of the course. If you want to better understand how to direct/prompt an AI agent, or a multitude of AI agents, to build applications then this is a really great foundation.

This was another excellent course from the ever-reliable, engaging, and much-appreciated Ed Donner. I learned a great deal and now feel much more comfortable working with agents—and perhaps even expressing my feelings to them. The only addition I would have loved to see in this already fantastic course is a deeper demonstration of the challenges that arise when using tools like Claude or Codex on large codebases, especially when they run into token limits while trying to complete meaningful tasks. Ed briefly touched on this, but in real-world projects, token exhaustion can become a significant productivity bottleneck. I would also have enjoyed a section on interacting with self-hosted LLMs via Ollama and using tools like Claude or OpenCode to generate and iterate on projects. Exploring the practical limitations and pain points of that workflow would have been incredibly valuable. Overall, it was a great course, and I genuinely appreciated Ed’s clear, approachable teaching style.