Microsoft embodies the cloud and all it encompasses including edge computing, AI and machine learning, mixed reality, autonomous systems and more. Microsoft is a good proxy for what tomorrow’s cloud-driven economy may look like. Our view? A tapestry of natural language AI-based applications and autonomous systems, all running on core platforms primarily powered by a combination of Microsoft Azure, AWS and GCP.

Microsoft as a Proxy: MaaP

Microsoft is perhaps the best proxy for the future of cloud-based software systems if for no other reason than Microsoft touches everything from the data center to edge computing to operating systems (although a laggard here behind Chrome and iOS) to embedded software, AI, machine learning and deep learning, as well as everything that Microsoft does at the application layer including Microsoft 365, the Open Data Initiative and LinkedIn. Azure is the underpinning for much of Microsoft’s strategy and is a strong number two competitor to AWS. Google’s competitive offering – GCP – boasts the strongest AI and machine learning chops and ought to gain share as the firm invests in supporting GCP’s partner network.

Below are my thoughts on some of the subjects that Microsoft CEO Satya Nadella covered during yesterday’s Microsoft Build conference keynote. A video of the full keynote is included at article’s end.

Conversational AI

• AI, Machine Learning, Conversational AI: We have come far, but it is still early days on the AI/ machine learning, deep learning front. We are years away from having a true cross-domain, natural language conversation with an AI. The amount of data required to train the requisite neural networks that would power a natural language conversational AI platform is enormous. Heck, the amount of data to train Google Assistant to schedule a reservation or Amazon Alexa to select a song is enormous. However, conversational applications that leverage Microsoft Bot Framework have gained traction. Some 3,000 conversational applications are created each week.
  • In our view AI and machine learning products will first be constructed to master specific domains. A good example of this is SS&C’s investment in AI, machine learning and RPA to automate investment operations. Over time, advancements in neural networks – the central nervous system for deep learning constructs – will enable conversational AI-based services to execute seamlessly across domains. However, this is years away.
  • Given the amount of data and resources required to teach neural networks, it will undoubtedly be Google, Microsoft and Amazon that lead the cross-domain, conversational AI investment effort. $100B or less market cap companies don’t have the scale or resources to compete.
  • We expect smaller technology companies to build cloud-based services on top of core platforms such as Azure, AWS and GCP. This essentially represents a continuation of the current trend.

Azure, Intelligent Cloud & Edge

• Azure/ Intelligent Cloud & Intelligent Edge: Azure is Microsoft’s cloud-based infrastructure (“the world’s computer” to quote Nadella, including 54 data centers globally), that powers Microsoft’s “Intelligent Cloud” and “Intelligent Edge” – two phrases that Microsoft’s management team and marketing machine have increasingly pushed in various communications. The former is the combination of Microsoft’s public data centers (“the cloud”) and core AI services which serve as the underpinning for a vast number of smart applications/ cloud-based services. The latter refers to edge computing or connected systems & devices where real-time analytics and services are delivered at the “edge of the network” – i.e. “close to users” where those insights and services may be delivered quickly, efficiently and securely as opposed to routing the workflow back through the central point of the network, chewing up bandwidth and causing latency.
  • Starbucks served as a keynote customer example as they have adopted more than 50 Azure services across the organization. These include: 1.) Starbucks’ in-house, AI/ML-powered platform – the aptly named “deep brew” – which incorporates an intelligent food and beverage ordering recommendation engine; 2.) Starbucks is leveraging Azure IoT to connect essential coffee equipment across 30,000 global store locations to Azure where predictive maintenance models may be run and from which coffee menu updates may be sent from Azure to coffee machines; to 3.) incorporating Azure blockchain into elements of Starbucks’ supply chain, enabling customers at store locations to scan and view the history of their purchased coffee beans (This video case study begins at the 1 hour 8 minute mark of the full keynote video at article’s end).
  • Microsoft has rolled out mixed reality services that combine elements of Azure and edge computing. Paccar is using mixed reality to deliver employee training services. Philips is leveraging mixed reality to execute non-invasive surgery and PTC is using mixed reality for industrial design (Learn more about these services at the 1 hour 25 minute mark of the below video).

Autonomous Systems

• Autonomous Systems: this is a new area for Microsoft that blends elements of simulation and “machine teaching” – automating the teaching process from the perspective of imbuing machines with domain knowledge such that they may execute semi-complex tasks. For example, Shell uses a Microsoft autonomous system for precision drilling. (Learn more at the 1 hour 27 minute mark of the below video).
• Open Data Initiative (in partnership with Adobe and SAP): Microsoft touched on its partnership with Adobe and SAP (two logical acquisition candidates for Microsoft) which you may read about here.
• Microsoft 365: We won’t cover in detail here. However, you may learn more about beginning at the 1 hour 29 minute mark in the below video.

If you are interested in watching the full Microsoft Build Keynote (aired live May 6 2019), check out the YouTube video below. Satya Nadella’s Keynote begins at the 54 minute mark.

• Software
• Strategy
• Technology