The Impact of Digital Transformation on the UC Experience
Matt Christopher, VP Customer Experience
The Next Wave of Traffic on the Network
As an opening to this blog, I was preparing an exhaustive exposé on the impact of sequential waves of immigrants throughout the history of the United States and their impact on the economy and on society. I was going to draw insightful and provocative parallels between the impacts of these immigrants on the US and the impact of new waves of applications on the network. In the end, that seemed a bit extensive for this particular blog. It also seemed like A LOT of work, so I’ll skip to the punch line.
From the time data networks were first introduced to the enterprise, there has always been a new “killer” app or technology wave that has promised to transform business productivity. Each of these have an impact on the network and IT infrastructure and each new wave believes they are the most important thing on the network. In the late 1990’s, that technology was Voice over IP (VoIP). VoIP evolved into Unified Communications (UC) which promised even more productivity, but took up even more bandwidth and was even more “important” than VoIP.
For the past decade, “cloud” has been a very popular buzzword/megatrend with “big data” entering the IT lexicon about 5 years ago. 2017’s technology trends included Artificial Intelligence, the Internet of Things, and voice search. In summary, over the past decade, networks have become more distributed, are transporting more data and are carrying more “transactions”, yet the user tolerance for degraded voice/video quality has not slackened.
The Promise of Artificial Intelligence
The Association for the Advancement of Artificial Intelligence defines Artificial Intelligence (AI) as “the scientific understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines.”. AI leverages learning, deep domain-specific knowledge, reasoning, and interaction to automate many functions and expedite the process of finding novel solutions to common problems. From Watson winning Jeopardy, to Alexa setting the thermostat, to personalized cures for cancer, the promise of AI ranges from the novel, to the practical, to the life-saving. In the world of enterprise IT, AI promises to automate many tasks and drive efficiencies throughout the overall IT infrastructure.
The (Hidden) Impact of Artificial Intelligence
What do machine learning, reasoning, data processing, and real-time interaction have in common? They are all key components of AI and they all consume resources on the network and within the data center. With the promise of AI comes an impact on the IT infrastructure and a potential consequence to other real-time communications applications like UC.
Preparing for AI
As AI continues to expand its reach across enterprises, network administrators must plan for the additional burden on the network. They must make sure the network can handle the extra load and that true real-time applications are given real-time priority.
There are 3 things IT professionals can do to protect the UC user experience against the propagation of AI-related traffic.
1. Prioritize Voice
Another buzzword of the past 5 years or so has been Software-Defined Networking (SDN). When applied properly, SDN technology, like Cisco’s APIC-EM, can be leveraged to make sure that voice and video are always tagged properly and given real-time priority over all other traffic on the network. UC-specific tools can leverage SDN technology to ensure the network is always aware of UC traffic and prioritizing it appropriately.
2. Optimize Network Functionality
As traffic over the network increases, the headroom for error goes away. It is important the IT staff optimize the network to properly route all traffic. In addition to network upgrades, configuration updates, and SDN, IT staff can utilize synthetic transactions to ensure the network can handle the load and still give voice and video the room they need to not impact the UC user experience.
3. Quickly Diagnose
Even with all of the planning and automation, there will still be instances of poor call quality. As AI becomes more popular, the expectations of human users will not disappear. If there is a bad call experience, the IT staff must be able to quickly identify where (if anywhere) on the network the congestion occurred and resolve it as quickly as possible. The addition of AI traffic just creates more “noise” on the network and makes root cause identification that much more difficult. IT or their managed service providers can leverage tools to quickly pinpoint the network chokepoint and resolve the bottleneck.
The Machines Aren’t Taking Over (yet)
AI has tremendous promise, but it still needs the IT staff to create an environment for it to flourish. This same IT staff must also prune the network to ensure that the promise of AI does not come at the cost of human-to-human interaction over the network.
Special thanks to Daniel Castro and Joshua New for their paper, “The Promise of Artificial Intelligence”.