Excellent

отличнно

Cool !

nice

good

good

Good

yes

да

Ok

а

You will need intermediate level Python coding experience (OOP in particular) to understand the topics, so it's not at the beginner level. Also the documents could be in a better order or written with more clarity. Other than that, the course shows you the ease with which AI agents can be created by mixing a LLM with a python coding framework that you can build from scratch. Quite amazing!

I really enjoyed this course by Prof. Jules White. It goes "behind the curtain" of how agentic AI is built through the "system", "user", and "assistant" prompt loop. One thing that would make the course 5 stars is if there were more graded assignments, including peer-graded homework.

Good examples - although they make you work a bit to really get them working.

its good for those students which are beginners in ai

awesome!

valued

good

The course presents useful concepts, but I encountered two inconveniences: 1. Although the course is marked as suitable for beginners familiar with Python, in module 3 there is a large AI agent project composed of an unstructured list of classes. It took time to analyze and draw a UML diagram to understand all the relationships in this core module of the AI agent. 2. For this course, I chose to use LLM models running locally on my machine via Ollama (models: Magicoder, GPT, LLaMA3). I ran the course code locally, and each time I faced issues during the agent loop execution—often the LLM failed to stay within the tool-based narrative set by the prompts. I don’t know how it would perform using the OpenAI API.

Lots of reading on your own and trying out the code on your own. Not what I expected.

Needs more diagrams

The core of the information is really good. But... for a "beginner" course, a LOT of python and programming knowledge is assumed. There are also very few graded assignments, which made it easy for me (with significant programming xp) to breeze through the course; but also didn't really solidify anything. I went with the recommended google.colab suggestion for running code; but it's not like I have files there. This would have benefitted from more programmer-esque setup and course structure (like, implement this part of the agent). More guidance on how to set up a local environment and and sample project to follow along with would have been really nice.

a lot of coding that i would have liked some of it explained in a video.

Muy pobre