If a company leader has to integrate AI into a large global company with tens of thousands of employees spread out in offices around the world, they’re likely to feel some discomfort.
Liz Centoni, Cisco’s chief customer experience officer, described the process as “surgery without drugs.”
“It’s painful,” she told Business Insider.
Cisco provides networking, security, and collaboration technology for enterprises, including routers, switches, cybersecurity tools, and services that help connect and secure enterprise IT systems. That’s why a customer support team is essential.
Centoni, who heads Cisco’s customer experience division, which has about 20,000 employees, said the company is currently transforming the division into an AI-native services organization.
Adopting AI requires more than simply incorporating it into old workflows, she said. This is what Cisco tried before realizing it only sped up a flawed process.
One example was customer support.
Cisco first used generative AI to create case summaries when one support engineer hands off a case to another for reasons such as shift changes, availability, or the need for different expertise. The goal was to provide more context to engineers. But Centoni said that while it streamlined handoffs, it didn’t address the larger problem and only “frustrated customers faster.”
“The result was never a handoff,” she said. It was, “How do I meet the right engineer from the beginning?”
This realization led Cisco to redesign its workflow to use “intelligent routing” to send cases to the right experts from the beginning. Cisco’s Centoni said the customer experience department handles about 1.5 million support cases a year, and nearly 88% are now routed to the right engineer the first time.
A Cisco spokesperson said the company currently measures customer service success by the number of calls that require only one handoff or even zero.
The best place to integrate AI is in repeatable workflows that can run autonomously with greater than 90% accuracy, Centoni said.
The company recently announced Cisco IQ, a digital interface for support and professional services. According to Centoni, it is “designed to be a single source of truth for customers to address recurring pain points.” He said this can help detect avoidable outages, redirect employees who spend too much time “interpreting data” rather than acting on insights, and reduce frustrating support calls.
For Centoni, testing an AI project is about more than just whether it improves efficiency. He said each initiative needs to show what work it’s stopping and how it will increase revenue, expand margins, deepen customer trust and enable the team to build next.