All the lab exercises are outdated. The llms provided by coursera are not compatible with the environment provided in coursera. Even the solution provided by coursera has errors sometimes. Due to this, I had to reprogram multiple parts of the practices, taking longer than the expected time (1h). As a result, I haven't been able to complete the practices as my schedule passed. I'm working while I learn this course and fixing the inconsistencies of the examples provided by the course just consumes too much time for me and I'm not being able to complete everything on time.

Mostly all the lessons are AI generated and the explanation looks simply bad.

Good overall but; - Some outdated knowledge (covers LangChain 0.2 while 1.0 is already out) - Labs are very wordy. - Labs estimated time to finish don't reflect actual time needed (one lab took me 3h instead of 1h).

Videos can be more intuitive for better engaging focus

Course is out to date. The theory is still valid, but most of the practical and labs is out to date with libraries and implementation with old version. Would be very helpful if before take the course we know the last programa, course and lesson updates to make a better choice. For example: Some course we can find "Recently updated! February 2026", but most of them we do not know.

The title should be "An intro tutorial in LangChain and Flask with basics on Prompt Engineering". Please get rid of the AI voice. It's really hard to understand and follow. The voice is very "forced" and clips the syllables.

The subject matter is relevant and worth learning, but the execution is poor. Labs have multiple bugs that were never caught before publishing — including broken starter code where custom parameters are silently ignored, making entire exercises non-functional. The lesson I'm on is listed as 45 minutes; I've spent several hours on it due to bugs alone. What makes this worse is that this is IBM. IBM built its name on precision, quality, and getting things right. This course is none of those things — it's lazy, untested, and beneath the IBM brand. They're offering a certificate at the end of this specialization. But ask yourself: would IBM actually hire someone whose qualification was this certificate? Based on the standard they've applied to building it, the answer is no. When I reported a previous issue, Coursera pointed at IBM and IBM pointed at Coursera. Neither prorated my subscription for the time I lost. If you're paying by the month, these bugs cost you real money. Proceed with that in mind.

WORST TEACHINGS HERE, NOT PROPERLY TAUGHT. EXPECTED MUCH BETTER .... I WANTED TO DOWNLOAD ALL THE PDF AND PPT SLIDES SHOWN IN THE VIDEOS. FOR SOME REASON I DONT FIND ANY ??? REALLY WORST TEACHERS HERE , THEY DONT PROVIDE PDF FILES FOR THE PRESENTATION. I WILL NOT RECOMMEND THIS COURSE TO ANYONE

Most of the content is poor quality, repetitive, surface-level, and entirely AI generated (both the script and audio). Also the lab doesn't work and nobody is responding to people complaining about it on the forums

Feels like an AI voice is just reading a summary for you, rather than a person teaching you

This course provides a clear and practical introduction to building generative AI applications. The examples are well-structured, and the quizzes reinforce the key concepts around models, prompting, LangChain, RAG, and fine-tuning. I especially appreciated how real-world scenarios were used to explain complex ideas in a simple way. A great starting point for anyone looking to get hands-on with generative AI.

IMO, the ungraded app items with the application explanations and examples are excellent learning and practicing tools, which provide the most straightforward way to get to the point and develop proper intuition and knowledge. Personally I spent quite a longer time with them, then the course schedule suggests - and this obviously required longer use of the corresponding shared resources.

I already have ang working experience with this, but this serves as a refresher and deep learning what's happening under the hood esp the concepts and implementation, making it more easier to work with AI using the correct terms, approach and methodoloties. I really appreciate this course especially the last two courses. Awesome!

Great intro course for folks with familiarity of the concepts of AI that starts to cover the interactions with the LLM using tools like FLASK instead of just covering basic concepts like prompting and AI basics. As the first in the series, i am excited to see what the rest of the course has to offer. Highly suggest.

Working through the provided lab is the key to getting the most out of this course.

Good introduction - and a smart ad for IBM.

Theory is good but code is outdated plz update that . also the chat model part like human message ai message none of code is working

I would say the course and especially the readings and questions are mainly AI generated. The topics are clearly layed out, but the questions are just randomly associated with those topics and do not contribute to any understanding.

Very high level - limited usabiliy - code examples in large part outdated as of 2026

Most of the contents are taught by an AI voice, which is very monotonous, same-paced and lacks any immersive experience. This is just a very fast-paced reading by some AI, not teaching something meaningfully. Waste of time!