Today at Build Microsoft Executive Vice President Harry Shum, who heads the AI & Research organization, talked about how AI is impacting society and how we are supporting developers to infuse AI into all their efforts. If you missed his keynote I encourage you to watch it on demand. He also talked about how AI will redefine broad sets of business processes across many, if not all, industries which in turn has helped me find my next mission within Microsoft.
The thing that I am most excited about with this year’s Build conversation is our ability to highlight how AI will impact developers and our ability to demonstrate both the platform to enable that work and how we’re using these capabilities to enhance our own products. I encourage you to try out the AI technologies in your apps and solutions, we think you’ll be surprised on what you can accomplish.
AI at Build 2017
For developers, AI introduces changes to how software will be designed moving forward. Where generations of developers have focused on programming everything that the computer did, future programs will get insights from the data and take predictive actions based on advanced algorithms, machine learning, deep neural networks (DNN) and a broad set of tools to continuously refine and optimize business processes and outcomes. During Harry’s Build keynote, he talked about some of the technologies within the Microsoft AI Platform that we are making available to assist developers. These include updates to the broadest collection of Cognitive Services with a new Video Indexer Service, the Cognitive Services Lab plus the ability to customize core cognitive services such as Speech and Vision. Building on last year’s conversational AI announcements, we have also made updates to the Microsoft Bot Framework to make developers more productive while developing bots, introduced adaptive cards for cross channel/platform bot development, the ability to publish to new channels and enabled the implementation of our payment request API. Developers can also build a skill for Cortana via the Bot Framework Channel announced today. These skills will run across Cortana apps and new devices including the new Harman Kardon Invoke speaker and the Cortana Skills Kit is now available in the US, with other regions coming. To learn more about how developers are using Cortana Skills check out our video. In addition, some developers will want to use compute in the cloud to run their own models. In Azure, we have made this easier by supporting the ability to run models on several CPUs, GPUs and eventually FPGAs. Harry has more details on many of these announcements in his blog post.
Redefining business with your data
Compute and algorithms are critical to having a powerful AI engine but data is the fuel that makes that engine run. At Microsoft, we are very fortunate to have a variety of unique data sets. Bing and LinkedIn provide a wealth of world knowledge, while the Microsoft Graph provides a wealth of work knowledge including signals from Office 365, Dynamics and even devices. When these data sources are combined with your data there is the potential for incredible solutions.
With cloud compute, data, AI platforms, tools and algorithms; developers have a variety of options for using AI within their products and solutions. Yet for many businesses, even with these building blocks there will be significant hurdles to overcome. AI expertise is in short supply compared to the significant demand. Business leaders often don’t recognize what AI is capable of, how to apply it to their business and most don’t have the AI expertise to integrate AI into their business processes.
As Harry mentioned, AI has the potential to disrupt every business app and process in every industry. Across industries and departments like sales and marketing, we are working with developers and businesses to avoid being disrupted by redefining their businesses with AI using our platform technologies, data, AI expertise and our long history in developing for enterprise needs.
We have been applying AI to our own business processes and needs with great initial success. For example, contacting customer support can be a difficult experience for many. Customers want solutions to their problems now in a personalized and convenient way. Agents are challenged to handle these broad range of questions across different channels, to provide the best solution for the customer. Microsoft’s Customer Support Services team uses our customer care intelligence solution to address this challenge. Applying ML and analytics to our knowledge bases, including FAQ documents, web pages and historical chat data, we extract the knowledge required to effectively answer our customers’ questions. Conversational AI helps us answer in a personalized way. Using our bot and AI technologies, we provide expedient self-help support to our customers or connect them to an agent augmented with the conversation context and AI-driven content to better serve them. To continuously evolve the system, we used deep reinforcement learning, to learn from each customer/agent interaction.
Another area of focus is helping our finance teams make the most informed decisions they can. People have often turned to Bing to help pick March Madness, World Cup and Oscar winners. Beyond sports, entertainment and politics, there are key business scenarios where powerful insights can impact a business’ bottom line. Microsoft’s treasury uses our market intelligence solutions to help manage our cash and short-term assets. Institutional investors leverage our macro-economic predictions to better inform their trades and investments. We apply our ML algorithms for entity extraction, sentiment analysis, query understanding and more on aggregated and anonymized data from Bing and the web to do forecasting, financial predictions, product reviews and more.
When world data is combined with workplace data and your business data, incredible insights and associated actions can be enabled. Combining various signals from marketing automation, CRM, Bing, LinkedIn and interaction patterns can help sales people quickly triage leads and accelerate closing deals. Computer vision can create safer work environments by identifying safety hazards before there is a problem.
There is a lot happening in this Business AI space and we are just getting started. We will have more to share in the future.
DX Leadership Changes
With the advancements in AI and the opportunity to redefine businesses with AI, I’m happy to use this BUILD to transition into AI & Research. Over the last 4.5 years I have had the good fortune to lead the developer evangelism group (DX) where we have worked with developers and ISVs from around the world to help them leverage the incredible advancements in software and tools to deliver incredible new products, solutions and tools for their customers and partners. This journey has included interacting with students exploring computer science, fast growing startups, enterprise developers, technical and business communities, ISVs and so many other interesting and thoughtful individuals. Over just the short time of this journey, software development has changed significantly but the drive of developers to shape the world through software continues. As Microsoft accelerates its AI investments to amplify human ingenuity, I look forward to seeing what Charlotte Yarkoni, CVP, John Shewchuk, Technical Fellow, and Ron Huddleston, CVP, do for developers and ISV’s as they onboard the teams that previously comprised DX.
It’s been a privilege running Build for so many years, I look forward to attending next year, and hope you enjoy the Day 3 keynote video from John Shewchuk and I on Channel 9 this Friday but you can see the exciting trailer here now.
I’m looking forward to seeing what you build!