Artificial intelligence (AI) has ushered in a new era of interaction insight to the unified communications (UC ) and contact center space. But only for those with the platform packing the brainpower to decipher the torrent of data that flows through their UC systems.
At its core AI is about data rather than ‘intelligence’ or the so-called ‘artificial consciousness’ that some claim is about to awaken and take over the world.
Mostly, AI is statistics. Statistics drive most human decisions. For example, when you encounter a pair of headlights coming at you in the darkness, you’re likely to think: “Well, experience tells me that it’s probably a car rather than bikers riding two abreast. So perhaps I’ll step aside rather than gamble on standing still in the expectation that the bikers will pass either side of me.” Such thinking applies statistics to decide where to stand when encountering a set of approaching headlights.
Voice recognition works in a similar way. For example, when you say ‘hello’, a sound wave is created – complete with certain attributes – and stored in a database. Every subsequent ‘hello’ (and, for that matter, all other words) is also recorded. Attributes of each sound wave undergo statistical analysis to confirm and identify the words.
Once recognized, the system links the word to code – a command – which, in the case of ‘hello’, says: “When someone says hello, say hello in return.” The system learns and improves voice recognition with each new ‘hello’. It’s smart, for sure, but it is statistics that does the hearing.
AI applied to UC and contact center management works in a similar way. But instead of words and sound waves, data and its attributes relate to UC system components and their performance. So, for example, AI can troubleshoot issues by drawing on past experiences to solve new and emerging problems, comparing attributes to detect common patterns and draw conclusions about the nature of new problems and how they should be solved. And when that happens in the cloud, where an AI-enabled UC management platform draws on incident data from thousands of customer deployments across the globe, learning is more advanced, simply because there’s so much more data.
Virsae Service Management (VSM) adopts this approach, collecting data on a massive scale, overlaying AI-based analytics to drive automation and workflows to keep the technology that underpins customer interactions running flawlessly.
The data our system analyses shows the potential for AI to smooth over significant kinks: 33% of businesses have call queue issues; 23% have suffered from shaky voice quality over the last three years, largely as a result of increased competition for network resources; 74% are using less than 50% trunk capacity; and 68% could benefit from right-sizing capacity, including trunking or license counts.
As it stands, VSM is putting AI to good use, in a typical month automatically remediating 38 million incidents, monitoring half a billion calls for user experience, and assessing one million channels for capacity utilization.
Vendors make it extremely simple to adopt UC. However, managers often underestimate the planning and attention required to keep voice functioning at peak performance, when it competes with other data and users sharing the same network.
Voice quality is frequently problematic – and commonly misdiagnosed when administrators don’t have service management tools built on AI. Instead they rely on network monitoring tools that serve up symptoms of more fundamental problems they’re unable to identify or resolve.
With the right set of eyes and a platform like Virsae’s VSM, UC managers avoid the pitfalls of short-sighted monitoring tools, keeping calls crystal clear. A holistic approach considers the entire mountain of data – not just the visible tip. And this is where AI does the heavy lifting, looking for patterns in far corners of the UC ecosystem to identify the problem and propose – and even automate – a solution based on earlier successes and failures.
AI shifts the focus from raw data to what it means and critical things to do. And when a service management system learns from problems encountered and resolved at other companies, it can proactively respond to similar circumstances elsewhere, making it artificially intelligent.