Tuesday, December 13, 2016

Microsoft brings its flavor of AI with Microsoft Cognitive Services

As major vendors look for ways to differentiate in the cloud, application development and artificial intelligence are two key tools in their arsenal.
Salesforce has long cultivated its Lightning development platform, and recently released Einstein, its artificial intelligence application. At Amazon re:Invent in November, the cloud provider announced a series of artificial intelligence (AI)-powered APIs for voice, text and video.
These announcements came on the heels of yet another flavor of AI, Microsoft Cognitive Services, which was released in March. The series of more than 20 APIs enlist AI for video, audio, text and more. SharePoint and BI expert Scott Robinson joined Search Content Management for a podcast on the importance of these services for users.
"They are Microsoft's way to capitalize on capabilities they have already built into their cloud environment," Robinson says. "They decided to make APIs available to the user base so they can create apps. [Microsoft Cognitive Services APIs include] face recognition, object recognition, face identification, OCR [optical character recognition], speech recognition, text [analysis] and sentiment analysis. All are based on a machine learning component, so you have AI you can build into any application."
Robinson says there are many different directions to go in, given the variation and volume of APIs. Two of the greatest benefits of these APIs, according to Robinson, are improvements to customer experience and bringing heft to processes like social media monitoring, which scours the web for comments that may be favorable or unfavorable to a company.

"People want to improve relationships with customers," Robinson says. "This is a big part of social media monitoring, and it cranks up the authenticity, veracity of social media monitoring by an order of magnitude."

Some experts have indicated that voice is the form factor of AI for the future, and Robinson agrees.

"Voice is prime-time ready; video is not," he says. "It's one thing to analyze a face. It's another to pull up a video square dance and analyze the video for information. We're moving more toward hands-free interaction."