CV_MayJune_2018

predicting if specific devices or user networks are going to cause quality issues on a call and setting up alerts and mitigations to that effect. “The goal is to take action based on predictions to prevent a specific outcome from happening,” Shen said. “Ideally though, we would want the software system to take action independently from a people-centric managed service. If we can get to that level, the power of AI takes it up a few notches and the perception of a self-healing environment starts look- ing like a potential reality.” Self-healing would not only drive a reduction in cost but also an in- creased speed of the reaction. “For example, if you’re driving an auto- mobile that has self-driving features, you don’t want it to warn you when there’s an object ahead leaving you to apply the brakes,” Shen explained. “You want it to detect the object ahead and automatically apply the brakes, ultimately saving time and avoiding accidents. That applies here for UC systems as well.” Early Days Despite the excitement, much about AI for UC is hype today, mainly because it’s an entirely different way of looking at analytics. “Existing UC analytics tools come from a heritage of deterministic program- ming – showing reports, showing certain anomalies within the reports, highlight- ing different things – but that can only get you a certain part of the way there,” said Shen. “Once you apply the po- tential of machine-learning algorithms, the complexities of the problem you’re able to solve goes up by 10 or 100-fold, because you’re no longer reliant on the programmers to code this in a determin- istic fashion. The hype is more just in the state of the art of the implementation. In the UC realm, AI is still experimental. There isn’t a guaranteed outcome, and, like any research project, you might have to run 10 experiments before you have an outcome, or you might have to run 100.” Of course, as the Microsoft, IBM and VMware investment demon- strates, there’s enormous interest in creating real solutions, with resulting intellectual, time and financial ben- efits. Zeus Kerravala, founder and principal analyst at ZK Research, said at a recent CIO summit that the potential is undeniable. “I think we’re really at the point of infancy for AI in the workplace; it’s something that’s been theorized by companies like Microsoft and Cisco for a while,” he said. “There are a lot of CIOs interested in it, and I think they’re still trying to figure out what to do with it. We’re at the very early stages of [AI in the workplace], but over the next couple years we’re go- ing to see tremendous innovation for AI in the workplace. It’s going to make us more productive.” And the place you’ll see it first, said Kerravala, “is in corporate meetings.” o Core Communications Consider KADENCE Circuit Management Fiber networks represent a substantial asset and once construction is complete the essential task of managing that network and generating an ROI ensues. KADENCE Circuit Management (KCM) has been enhanced with functionality specific to the management of optical equipment, mapping of fiber assets and the tracking of outside plant (OSP). gokadence.com 866-877-4373 Customizable Workflows Maintenance Notifications Manage SLAs via NMS Power + Space Management Contract Management Network Discovery SLA MSA Your fiber construction project is complete… now what? May - June, 2018 | Channel Vision 21