the Cloud

Microsoft’s Reorganization Is All About AI and IoT

At the end of last month, Microsoft’s CEO Satya Nadella stunned many industry observers by announcing a reorganization of the company that would shift its focus away from the Windows operating system that had given it a near monopoly on the personal computing industry. The head of the Windows division, Terry Myerson, was stepping down (or was asked to step down) and the division was to be split up into the two new organizational units: Experiences & Devices and Cloud + AI.

Since then, Microsoft has announced that it will spend $5 billion on IoT research, products, services and programs. Much of this will be put into building its Azure platform. Side note: for those of you like me that were unsure how to pronounce Azure, it seems like there is a bit of a divide between the U.S. and the U.K. with the former pronouncing it with the emphasis on the first syllable (AZH-uhr) and the latter with it on the second (az-YOOr). Think of Azure as part AWS cloud hosting, part Apple AppStore, and part data marketplace. Microsoft, along with most other tech giants, wants to be able to sell artificial intelligence analytics much in the same way we purchase cloud data storage. To do this they have developed what they call the “hybrid cloud” which allows applications and data to connect both in the cloud and on-premises.

Designing the system with the capability to compute at the data source (or at the edge) is paramount to being able to enable AI applications like computer vision. Since the amount of data that needs to get processed for computer vision is huge the bandwidth capability of the transmission lines to and from the sensors becomes the bottleneck. A partnership was just announced between Microsoft and YI Technologies, an imaging specialist, that would put all of YI’s customers on the Microsoft cloud and potentially propel Microsoft to the forefront of the computer vision vertical.

Computer vision, which aims to give computers a visual understanding of the world, could be particularly useful in commercial real estate, an industry defined by maps, floor plans and other visual data. As Reonomy CEO Richard Sarkis, points out machine learning algorithms are already being utilized to process immense volumes of real estate data, streamlining much of the administrative legwork necessary to conduct market research.

Microsoft’s reorganization goes against what has traditionally been the company’s priority, controlling the operating system. Ex-CEO Steve Ballmer has said, “Nothing is more important at Microsoft than Windows.” But the stagnation of the PC industry in favor of tablets and phones along with customers’ desire to be able to have a “cross-device experience” has torn down the walled garden that their organization one sat behind. Now, they see the future of technology as enabling people and businesses to create customized systems rather than trying to keep them using their software.

Data storage went from a distributed model to a collective one in the cloud and if Microsoft gets their way, the same could be happening for advanced computing and analytics in the near future. Whether they are able to establish themselves as the leader of this new front isn’t certain but the clear winner will be the end user. Those trying to create advanced IoT networks will now have more options for designing systems and analyzing data than ever before.