What I Learned at Microsoft Ignite

By:  Michael Miller

At last week’s Microsoft Ignite conference, I heard a lot about Open Data; even more about security, including both tools and operations; and much about how Office 365 is evolving.

But there were also some broader themes I found interesting.

AI is Everywhere, and Being Used More Than You Think

It seems like AI has been the focus of every trade show and every vendor conference this year, and Ignite was no exception. And, like many vendors, one of the big trends at Ignite was the “democratization of AI.”

At Ignite, a big theme was “Everyday AI.” It was visible in the new tools in Office, such as the Designer tools in PowerPoint, which suggest new looks for your presentation; the ideas in Excel that suggest different chart formats; and the Focused Inbox feature within Outlook.

The effort is mostly aimed at developers. Dave Forstrom, Microsoft’s Senior Director of Communications for AI, told me that 1.2 million developers have used one or more of Microsoft’s cognitive services and 340,000 developers have used bot services. One big focus is making these tools available to citizen developers, so that they too can enable AI features, be these connections to services such as natural language processing or access to an organization’s “knowledge graph” with only a few lines of code.

Microsoft has a particular focus on chat bots reached via social media, such as Xiaoice, which Forstrom said had 200 million users in China, or Zo, which is in preview in the US. These conversational bots are “semantic machines,” and while the technology is not fully rolled out, Forstrom said 25 to 30 customers have actually launched customer-facing bots today.

Further up the spectrum is Azure ML, and in this area there is a push toward automated machine learning that aims to identify the most efficient algorithms and to optimize models, including tuning some of the labels. Microsoft has emphasized openness in AI, and to that end it is working with Facebook and Amazon Web Services on the Open Neural Network Exchange (ONNX) ecosystem. This includes new hardware acceleration for FPGAs, and a Python SDK for the service, which a number of developers have requested.

Data Is Still the Most Important Thing

In session after session, it was clear that when it comes to AI and analytics, having the right data, in the right format, is crucial. This is not only true for machine learning, but also for more traditional kinds of analytics. While many of the attendees I spoke with were excited about AI and machine learning, the vast majority were just as concerned about more traditional analytics.

Microsoft announced a variety of new tools, including a preview of SQL Server 2019—which is focused on building-in “big data”—and versions of Spark and the Hadoop Distributed File System packaged together, as well as connectors to other databases. The idea seems to be to turn SQL Server into a data warehouse for multiple kinds of projects.

Azure Data Explorer, also in preview, is an indexing and query system designed to help users quickly discover larger event data. Additionally, the Cosmos DB distributed database has been improved with multi-master support, a Cassandra API, and reserved capacity.

Of course, there are many other databases available as well, in both cloud services and on Microsoft’s platform. I’m impressed by some of the capabilities in Snowflake, a data warehouse product that started on Amazon Web Services and has since been generally released on Azure. Snowflake CEO Bob Muglia, who used to run Microsoft’s Server and Tools division, talked about how Snowflake has a different architecture compared to other cloud databases, and emphasized that this enables “multi-cluster shared data,” so customers can securely share information with other customers.

DynamicsFocus, LLC. is the leading source for accomplished Microsoft Dynamics, NetSuite and Salesforce professionals throughout North America, learn more here

How AI and IoT Work Together Is Important

It’s not a new idea, but I heard several speakers talk about how AI is increasingly being used in edge or Internet of Things (IoT) applications to make devices smarter. In his keynote, Microsoft CEO Satya Nadella called out one such application, noting how agricultural machine maker Buhler is using computer vision to look for toxins in the food supply.

Buhler Chief Digital Officer Stuart Bashford later explained to me how his company has been working on various projects to help make the food supply safer and more efficient. The company brought its new LumoVision grain sorter to the show, in which sensors rapidly analyze grains of corn and remove those that have aflatoxin. This machine uses a hyperspectral camera to obtain a 3D view of the kernels and Microsoft’s ML Studio for the processing; the whole operation runs on internal FPGAs and DSPs to make a decision within 80 microseconds.

Buhler is currently working with Whitworth Brothers, a big U.K. miller, on a blockchain solution to track grain from the farm to the store. It sounds similar to a blockchain initiative promoted by Walmart and IBM, but Bashford said this is different in that it isn’t controlled by a single store and is designed to work with very inexpensive commodities, such as wheat.

Microsoft announced a number of new features in this area. Among the most interesting were new tools for creating “digital twins”—or digital replicas of spaces and infrastructure—using the cloud, AI, and IoT to help reduce the need for maintenance and energy consumption, and the Azure Data Box Edge appliance, designed to make it easier to capture and transform data before it gets to the cloud. Azure Sphere, Microsoft’s edge security device, which had been announced, is now in public preview.

The Surface Hub is a Rotating, Multi-User Electronic Whiteboard

On the hardware front, Microsoft teased forthcoming versions of its Surface Hub computer, designed to be used as a big whiteboard (with video) for group collaboration in Teams.

In fact, the firm showed two different devices. The more basic Surface Hub 2S is due to ship in the second quarter of 2018, and is considerably sleeker than the existing unit. It has a 50.5-inch 4K display, a better camera, and is light enough so that it can be moved easily (and indeed, Steelcase has a stand for just that purpose). New features include one-touch log-in, so you can log into a meeting like you would a computer, and a number of improvements to the whiteboard feature.

The more interesting machine is the 2X, which was first demonstrated in the Ignite keynote and which has multi-user support so that multiple people can log into the same screen and each share their files, or etc. The 2X will also support tiling and rotation, and Microsoft demoed how the text on the screen stays in the right position as you rotate the machine from horizontal to vertical. The new Surface Hub line includes a computer within a cartridge that slides into the back of the screen, so you will be able to easily upgrade from the 2S to the 2X. Even though Microsoft wasn’t ready to share hardware specifications, these are the highest-end whiteboards that I’ve seen, and they look great.

The company saved its introductions of the new Surface Studio 2, Surface Pro 6, Surface Laptop 2, and Surface Headphones for this week. These are nice-looking devices, and Microsoft has done a good job integrating them with Windows, though there are a number of great-looking Windows laptops out there. Still, there’s nothing else quite like the Studio, which has an amazing screen, the Dial hardware, and the ability to lay flat. It’s pricey, but I can certainly imagine that it would be a big hit in design departments.

Microsoft Is Serious About Teams, and Many Partners Agree – But It Still Has a Way to Go

Microsoft spent a lot of time talking about Teams$8.25 at Microsoft, its latest collaboration platform, which currently has many of the chat features of a product like Slack$8.00 at Slack, along with connections to much of the rest of the Microsoft Office 365 suite. Many of the talks demonstrated how Microsoft is working to integrate more features into Teams, such as making Yammer, an earlier chat tool, a tab within teams, or linking Teams and Skype for Business$2.00 at Microsoft.

On the show floor, I saw I lot of companies pushing Teams integration. Crestron, which has long been known for devices for integrated media in conference rooms and similar locations, now offers a more traditional phone headset that’s designed for Teams, with software control hosted in Azure.

Plantronics had a dock that turns your mobile phone into a more traditional phone, also for use with Teams.

It’s not all that sexy, but it is useful. The concept behind Teams makes sense, and Microsoft seems pretty committed to the idea, but there are still a number of missing (but promised) features to come, so we’ll have to see how well it really works in practice. I have seen Microsoft talk about unified communications for a long time—I remember products like the Office Communications Server, Lync, and Skype for Business—and am a bit skeptical.

Ecosystems Win

The need for platforms and ecosystems has been a key part of the computer industry since the heyday of Windows (and arguably going back to the IBM 360), but it’s become even more important in an era when most companies want a “360-degree view” of their customers. To that end, Microsoft announced an Open Data Initiative, and Nadella brought up SAP CEO Bill McDermott and Adobe Systems CEO Shantanu Narayen. The idea is to share information, primarily relating to the customer experience, using the principal that each organization should have full control over their own data; the hope is to break down silos.

Of course, other companies understand the importance of alliances as well. During its conference the same week, Salesforce announced a partnership with Amazon Web Services to simplify how customers can share and synchronize data, and to link multiple services together. Of course, there are lots of other partnerships, as seen on the show floors at all of these kinds of events (not only Ignite and Dreamforce, but also AWS RE-Invent, Google Cloud Platform, Oracle Open World, IBM Think, and others). It’s clear that no one company will have all the answers, or all the data, and working together is, and will be, a key part of the technology that everyone needs today—and tomorrow.