AI: Leveraging increasing amounts of data to improve decision making

This past week, Accedian attended the Telco AI Summit and Software Defined Operations event in London. I participated in a panel discussion on “Service Assurance to Support 5G” moderated by Heavy Reading, along with colleagues from Orange, BT and Comarch.

The Telco AI Summit and Software Defined Operations was a broad event that included discussions around the applicability of AI and software operations from across the telco business spectrum – CRM, security, OSS/BSS and, of course, performance and service assurance. Everyone is looking to leverage increasing amounts of data to make more intelligent decisions.

Some clear trends are emerging that impact service assurance:

• Data and telemetry – There are more and more sources of telemetry data. We’ve talked about telemetry in this blog before. We continue to see more and more sources of data from the network – for example our partners at Intel were at event in London, showing off how you can pull telemetry from compute infrastructure. Another important aspect of AI is the utilization of clean and high quality data. Without quality data that is highly relevant to the answer you are trying to find with artificial intelligence for improved decision making, the answer, of course, would be quite hard to find. To read more about the importance of granular, high quality and clean data, read this blog by our Senior Director of Research and Development for our analytics solutions.
• Software and cloud – Speaking of compute infrastructure, another emphasis was the importance of cloud. From sessions at the event on multi-access edge compute (MEC) and migration of workloads to public cloud, a cloud-based service overlay will be critical to 5G service delivery.
• Dynamic environments – In the past the service assurance environment was largely static – network infrastructure deployed and monitored. Now we’re facing an environment that is much more dynamic; as use-cases emerge to be able to create and adjust infrastructure on the fly (e.g. network slicing) then the ability to deploy and manage service assurance into that same infrastructure will be critical as well.

The convergence of these trends is leading to a seismic shift in service assurance. Telemetry via open APIs, dynamic software-based test agents that can be deployed and reconfigured in real time and analytics solutions that can leverage AI to make sense of that data are all critical to being able to assure services as we go through the evolution to 5G.