Deploy a Production Machine Learning model with AWS & React (Udemy.com)

Clear and concise logic. I understood the AWS process: create an account, set IAM permissions, use SageMaker to download the dataset, upload to S3, train the model, and deploy the endpoint to Lambda, using API Gateway and CloudWatch along the way. However, the course is outdated (2–3 years). AWS is fine as the interface hasn’t changed much, but the React project struggled with Node.js updates. Fixing frontend and backend issues took time, and the framework is outdated.

This course is absolutely worth it! It gives a great overview of the tools in AWS to get you started in cloud ML. Teaching the necessary tools like Lambda, API Gateway, CloudWatch, Endpoints, and more in an easy to understand way. It's amazing how much valuable information was packed into this compact amount of time and it's only understandable that the webapp sections were more of an overview. Maybe in future there could be a course just the other way around, focusing more on the deployment part for ML and DS? Thank you Patrik for this wonderful course!

I have to say, it was an excellent experience! From start to finish, the course was well-organized, engaging, and informative. The videos were all professionally produced and high-quality, making it easy to follow along and stay engaged throughout the course. Overall, I would highly recommend this course to anyone looking to learn about the subject matter.

Great course. One of the few ones (that I've taken) covering not just the basics of model creation, training, validation but also how to deploy and use it in real-life scenario. It showcases the applicability of AI in the medical domain. Explanation was accurate and easy to follow - I 'walked' into the course mainly for getting more familiar with Sagemaker, I have to admit that I'm walking away with some useful ideas for front-end and back-end security as well. Well done! Some suggestions to further improvements to the quality of your courses/contents: - given that the model is trained on medical images, it would be preferred to use DICOM images instead of jpeg as they have diagnostic quality preserving all details captured by the imaging modality. - would be interesting to have a sequel to this course showing how to create custom models with user-defined layers of CNN. If it was done on a 3D volume, it would be even better. - the course focuses on classification. As a next step, I'm sure it would be welcomed if you manage to cover image segmentation topic also on medical images (can be lungs or other organs). Regardless, I very much enjoyed your course. I hope the above suggestions can help you on your journey and I encourage you to create more content!

Patrik explains things very clearly and patiently, even dwelling into the nitty gritties such as why we do certain things or type certain code. Much recommended! I could follow and understand even though I've no prior working experience with AWS

This is a really useful course, if you want to get more familiar with the aws tools in data science. The end of the course is getting advanced. So I don't really recommend it to someone, who is not comfortable with error handleing. (there are some parts of the course which is outdated, and therefore this will be a lot of work, to achive the same results. This is the deployment only, the aws part is up to date ( 2023.02))

This was a good course. I learned a lot about AWS, which was my goal. The instructor is very easy to understand. The only suggestion I have for improvement is to add one more lesson to the final section of the course so students can get the final end-to-end web site (to AWS and back) working. There was some missing information.

This was an absolutely amazing course and really look forward to others from him in the future. Up until now, most of the courses you find out there are focused on a scholarly implementation but then lack on the efforts necessary to productionalize your models. This course exceeded my expectations and I would recommend it for anyone looking at implementing their models in production. I also appreciated him taking the time to tend to the "best practices" from a security standpoint, as these are crucial when implementing your model in the real world. If there was one topic I would have loved to see him touch on, it would have been "how to retrain your model on new data in a production flow". Over time there may be new data that you would like to add to the dataset to increase the accuracy of the model and I would love to see something around this from a pipeline perspective in the future. Perhaps for a future course. That said, this was hands down one of my favorite courses on the topic and he was clear, to the point and I found the course to be well structured. I'll be looking forward to future releases from Patrik.

An amazing course on AWS and Machine Learning. Learning loads while making something that's relevant to industry, which will be great to show on a CV. The instructor clearly knows his stuff and explains it very well. The only downside is that this is his only course. This is by far the best Udemy course I've ever taken!

A relatively compact course with a straight line taking you from the data (in this case, images to be classified) up to a deployed model, with attention to details such as security and costs. I did not like 'Creating our web application' (because it is not about creating a web application at all, just a brief walkthrough of an existing app made by Patrik - not really surprising because making an application is enough for a whole separate course. Unless you have some reasonable knowledge of Node.js, React etc, you will probably find it difficult to follow what happens...) and I would strongly prefer making of a very basic app (simple enough for everybody to be able to follow) to this. Having said that, I did not come for making a web application but for deployment of a ML model and that's what most of the course is about, well explained and definitely useful.