The landscape of AI and ML technologies provided today by Microsoft is getting bigger and bigger and more complex every day. Cognitive Services, Azure Machine Learning, Azure Databricks, ML.NET, SQL Server, Power BI – to name just a few – are all options when it comes to developing and/or integrating intelligence into modern applications. With many capabilities overlapping in features it becomes increasingly difficult to decide on which technology or set of technologies is best suited for a solution, product, or scenario. Understanding in depth Microsoft’s galaxy of AI and ML offerings is not easy, to say the least.
This section is my first attempt to take a structured and methodical approach on analyzing the entire range of choices available in Microsoft’s technology landscape. This page is the central hub where I'll provide links to all relevant blog posts, conference sessions, and slide decks I am planning to use when speaking about Microsoft's AI and ML strategy and approach. If you're either in your early data science days or a developer looking to get into the world of AI/ML, this is the right place to find some guidance to bootstrap your understanding of Microsoft's AI/ML landscape and the architectural + technological choices you'll have to face sooner or later. If your're a seasoned data scientist looking to materialize your knowledge via Microsoft Azure, I hope you'll find the structured approach useful to make you proficient in record time. If you're a seasoned data scientist and a seasoned Azure professional, all I can say is thank you in advance for any kind of feedback related to the material :)
Either way, welcome to the ride!
The first step into the world of Microsoft AI and ML services consists of understanding the high level spectrum of offerings. It ranges from pre-trained, readily available models (trained on data provided by Microsoft) all the way to 100% custom models, designed, built, trained, evaluated, and operated by developers, sysadmins, or data scientists. The Microsoft AI/ML Spectrum post talks in more detail about this overview.
Azure Cognitive Services provide the left side of the spectrum, the pre-built AI and ML services. The capabilities of these services range from fully trained models to models whose training can be customized using specific, domain-related data. All models are exposed as REST APIs as well as containers that can be deployed on a wide range of environments. Read more about them in the Microsoft AI/ML - The Pre-Built Part of the Spectrum post.