DTW 2022: Verizon sees into the future with help from data and AI

Verizon has spent the last four years evolving its analytics practice from “one that’s more diagnostic and descriptive to one that’s more predictive and prescriptive” through the implementation of artificial intelligence (AI), among other factors.

Speaking at Digital Transformation World 2022, Shankar Arumugavelu, chief digital and information officer at Verizon, said the US operator “wanted to get better at predicting what the future can be and also provide guidance on what to do next. So that is where we saw the promise of AI.”

Data “is the raw material for this,” he added. “And as communication service providers, we don’t have paucity of data. We have tonnes of data.”

Verizon also believes that data “is only powerful in the presence of more data, and this is really where we set out to build a real time, stateful digital twin of all our assets,” Arumugavelu said.

This is fundamentally built on three principles, he explained. “How do we get more data about our assets. How do we get them at an increased frequency more near real time. And third, how do we build a common data fabric that can stitch the data across these different domains.”

He noted that the goal is ultimately to “create a customer digital twin, so we can provide that hyper-personalized experience for our customers.”

Arumugavelu provided some examples of where this new analytics practice is being put to work. For instance, Verizon applied computer vision to determine where to place 5G masts in dense urban markets “to get the maximum coverage with minimal number of nodes.”

The operator also monitors the total cost of ownership of running a site. Here, energy usage is of course increasingly critical. Arumugavelu said Verizon has been able to reduce energy costs by over $100 million a year by building energy consumption models for the digital twins.

In other examples, Verizon has moved away from data-driven workforce planning to AI-driven workforce planning, enabling it to make better overall use of its workforce. “For example, in a retail store, we would have our agents do lead targeting during those periods when that is lighter foot traffic,” he said.

Arumugavelu further highlighted churn models that allow the operator to more efficiently target its customer base and improve revenue generation.

“This is our AI industrialization blueprint,” he said. “Our ultimate goal here is around taking data, converting that to an insight and … acting on that insight. The whole objective here is to accelerate the time to insight.”

Good data governance and adherence to responsible AT practices are of course paramount. Overall, Arumugavelu concludes that AI is “no longer an emerging technology that can deliver value in the future. It is here now. It is making an impact now.”