Studying for the AI-900 exam


Recently released within the Microsoft Certification family was the AI-900: Azure AI Fundamentals exam. At time of writing this blog post, the exam is still in Beta format, meaning if you sit the exam you won't get your results back instantly. If you've never sat a Beta exam before or looking for some advice around it there is a great post here that covers off more information.

Now Artificial Intelligence, Machine Learning, is absolutely not my area of expertise, I am very much a infrastructure gal. However, I love to learn new things and stretch myself every now and again, and sitting the AI-900 exam and passing it is firmly in my sights, and I thought I'd share some of the resources I have bookmarked and am using for my study. I've based what I am studying off the official skills measured section on the AI-900: Azure AI Fundamentals site.

Microsoft's AI Principles

Microsoft has a set or principles that it follows when developing AI, you can find those principles here. I don't see this called out in the exam objectives specifically but I want to make you aware of what Microsoft's AI principles are as I think they are important when looking to what Microsoft is doing with AI and guide you as well if you are just starting your AI journey.

Getting Start with AI on Azure

Now I've dabbled a little with the AI products on Azure over the years, but never got in depth with any of the products and it has been awhile since I looked at them so this Microsoft Learn module is a great starting point for either learning about the AI products or refreshing yourself on what is available.

Natural Language Processing on Azure

A section of the exam 15-20% of it is around the national language processing or NLP workload being processed on Azure, and thankfully there is a Microsoft Learn module that helps you skill up on that particular portion.

Machine Learning on Azure

Now the machine part of the exam is quite a large portion, it accounts for 30-35% of the exam content. So this an area of importance, you definately don't want to skimp on knowing this area. My starting point is another Microsoft Learn module (can you tell I like Microsoft Learn 😉), the Create no code predictive models with Azure Machine Learning is a great place to start as it's a beginner module.

I've also turned to some of the Microsoft Docs pages for this area as well, this page has a great overview to an introduction of machine learning algorithms and there are some great visuals as well to break up the learning.

Conversational AI workloads on Azure

Conversational AI helps dialog between an AI agent and a human. The short Explore Conversational AI Learn Module can help you understand and get to grips with the basics of those workloads in Azure.

Computer Vision workloads on Azure

Computer Vision is all about trying to interpret the world visually using data sources such as camera, images, videos etc. Computer Vision pluse cognitive services can add an extra dimension. I'm currently working my way through the Microsoft Learn learning path Explore computer vision in Microsoft Azure, which is quite in depth in terms of what it covers, as it covers off computer vision, custom vision, face service and form recongizer.


Now this is just a start for the study material and I would say have a look at the exam objectives, the material I've suggested and do your own research. I am sure there is other great reference material out there that could help you study for the exam. If you have any suggestions please do reach out and let me know. 😀