This week, Microsoft launched a preview of its Azure Bot Service. Billed as a “bot service that scales on command,” the platform integrates with Microsoft’s templates for building bots, as well as the ability to quick scale up (or down) in response to demand.
Microsoft is following an industry trend that has seen Facebook and other tech firms release some type of bot platform over the past several months. Facebook’s bot initiative, which kicked off in April, allows developers to deploy bots for Messenger, its messaging platform. Whether associated with Slack, Kik, Facebook, Microsoft, or WeChat, many of these early bots are designed to automate either content delivery (weather, news, etc.) or customer service (answering questions and troubleshooting product issues).
Microsoft is betting it can stand out from that growing pack in two ways: by offering a variety of platforms for bot usage (including Skype and Teams, the latter a Slack competitor), and by allowing developers to leverage the computing power of Azure, including its machine learning APIs. The bots will supposedly interoperate on third-party platforms such as Facebook Messenger, Slack, Kik, and personal Websites (via a REST API).
“We want every developer to be able to build bots as the new application for every business and every service,” Microsoft CEO Satya Nadella told the audience at the Build developer conference in March, when the company first unveiled its interest in the technology. Developers, he added, should have the ability to “infuse” intelligence into the bots, opening them to respond to a greater variety of situations.
At that event, Microsoft unveiled the Microsoft Bot Framework, which includes SDKs for bots (in C# and Node.js) and tools (Bot Framework Emulator for Mac and Windows, along with a Visual Studio Template in C#). Also check out Microsoft’s BotBuilder on GitHub:
In a corporate blog posting Nov. 15, Lili Cheng, a distinguished engineer in Microsoft’s Artificial Intelligence and Research Group, again highlighted the Azure Bot Service’s reliance on the Microsoft Bot Framework, in addition to Azure’s “serverless” (re: cloud) environment. “Integrated continuous deployment support means that you can use your preferred DevOps tool chain and commit code to source control systems such as Visual Studio Team System, GitHub, and BitBucket,” she wrote, “and then automatically publish the code into Azure to continuously improve your bot.”
Those interested in what Microsoft has in store for bots can check out the company’s dedicated Azure Bot Service page.