Verizon is using AI to improve call center performance

  • Verizon told us how they are deploying AI in their call centers
  • A VP said that the AI helps staff learn how to best deal with customer calls
  • Apps include a live call transcription and a virtual assistant

MOBILE WORLD CONGRESS LAS VEGAS – When you hear GenAI, you likely think of ChatGPT. But viewing the developing technology through such a narrow lens isn't necessarily something operators should be doing. After all, GenAI is about more than just chatbots, a Verizon exec told Fierce this week.

Pranay Bajpai, VP of digital and assisted experience systems at Verizon, spoke with Fierce about how Verizon is using artificial intelligence (AI). He said that Verizon is first bringing AI systems to its call centers rather than rolling out GenAI chatbots. 

These call centers are run out of the Philippines, India, South Africa and Romania and tend to “have a lot of attrition,” he noted. That's a problem because it means Verizon loses a lot of institutional memory about call center processes when the staff move on. 

However, AI can help the call center manage and retain best practices about how the operator transcribes, routes and handles calls, Bajpai said.

Just using AI to route a call to the best customer agent to handle it has improved call resolution “drastically,” Bajpai said. Additionally, accurate live transcription of calls and what Bajpai called a “personal research assistant” means that employees don’t have to record or search around for particular data while on the line. The AI assistant delivers necessary data to them as they handle the call.

“It’s not about dollar value,” Bajpai said of Verizon’s initial AI efforts. It’s more about getting more productivity and improving staff performance as they deal with the undoubted multitude of customer calls, he said.

“I call ROAI,” Bajpai joked, combining the acronyms for return on investment (ROI) and AI. 

Verizon's approach is certainly one way to deploy customer-facing AI, an analyst told us.

"It certainly makes sense that operators start with lower risk applications such as Level 1 call center support,” neXt Curve analyst Leonard Lee commented in an email to Fierce. “GenAI augmentation of existing IVRs and chatbots make sense as initial adoption points."

“As we have seen for almost two years, adoption of GenAI in network management is tepid and cautious. We are seeing measured exploration in network security as well to determine where and how GenAI will be an effective and economically viable tool in network management,” Lee concluded.