AI-Embedded Service Operations Turn Contact Centers Into Proactive Experience Hubs

Most contact centers still run on a model designed around inbound volume: calls are answered, routed, handled, and measured on speed and resolution. That model isn't built for the expectation customers carry in 2026, which is that the organization should already know who they are, remember what happened last time, anticipate what they need next, and deliver service across whatever channel they prefer without making them repeat themselves. The organizations moving toward that standard are rebuilding their operations around AI as the orchestration layer, with human agents freed to handle the complex, sensitive, and high-trust interactions that define whether a customer stays.

Mitch Mann is Vice President of Member Services at VYTLOne, an independent provider of integrated pharmacy solutions, where he leads the contact center and member experience function. He's spent more than 30 years in contact center and CX leadership across American Express, Sprint, TSYS, Express Scripts, and more, with a consistent focus on building people-first cultures inside operationally demanding environments. That career-long vantage point shapes his belief that AI could be the most consequential capability shift the contact center has ever absorbed, but only if it serves the human experience rather than replacing it.

"AI shouldn't be layered onto the customer experience. It has to be embedded from the very beginning," he says. The embedding he describes starts at the first interaction and carries context through every channel and handoff the customer touches.

Proactive service is the new competitive standard

The clearest signal of the shift is that the best customer experience is increasingly the one the customer never had to initiate. Mann frames the transition around a principle he encountered in his research on next-generation CX. "It's about resolving issues before a member has to call us," he says. "'I see your refill is almost up,' or making outreach on a complex healthcare situation like prior authorizations."

That shift from reactive to proactive changes the fundamental operating posture of the contact center. Instead of managing inbound queues, the team is planning to orchestrate real-time actions based on what the system knows about each customer's history, preferences, and upcoming needs. "It's really not a static journey. It's about real-time orchestration where you're predicting the next best move," he explains.

Mann is clear that in order to be effective, the orchestration has to carry context across every channel. When a customer moves from IVR to portal to agent, the conversation cannot restart. "There's no repetition across channels as it moves from one channel to another. It's one continuous conversation. I can't get that call from the IVR to an agent and have to start all over again."

Humans handle the moments that build trust

The AI layer handles scale, speed, and consistency on routine transactions. The human layer handles the moments that determine whether a customer trusts the organization. In Mann's view, this is about changing agents' work rather than reducing headcount. "The future of AI is not about replacing people, it's about elevating people," he asserts. "There won't necessarily be fewer agents, but different agents. AI can handle scale so our people can deliver the moments that build trust."

The human-in-the-loop model he describes is critical to healthcare, where clinical and regulatory judgment cannot be delegated to automation. A medication override, for example, might surface during an AI-assisted self-service interaction. "That's something we'd want a member advocate to look at. But we don't have to transfer that call. The AI can put that interaction into a queue in real time. A human can look at it, make a decision, and it goes back."

Call note summarization is one of the first capabilities Mann's team is implementing. The benefit is dual: agents stay present during the conversation instead of splitting attention with note-taking, and the documentation that results is more dependable. "Not only is it going to save time, but it's going to drive much more consistent note taking."

Metrics have to change with the model

The operating metrics that defined the traditional contact center, like average handle time, call volume, and after-call work, do not capture the value of proactive, orchestrated service. As a result, Mann calls for a measurement framework that evolves alongside the operating model. "We're going to have to measure things differently. The old contact center metrics are important for forecasting, but we're going to have to truly figure out how we measure customer effort, journey completion, end-to-end resolution, and how effective AI is, not only with resolving the call but also member satisfaction with both AI and a human agent."

The comparison between AI-handled and human-handled satisfaction scores becomes a new diagnostic tool. If AI-assisted interactions consistently score lower on satisfaction, the system needs tuning. If they score comparably on routine transactions, the case for shifting human effort toward complex interactions strengthens.

Shifting from support function to experience hub

The structural implication Mann sees is that the contact center stops existing as a siloed support function and becomes embedded into the broader operating model alongside account management, sales, and product. "Member Services has to move from being just a support function to one that's a primary driver of retention, satisfaction, and value. But it has to start with organizational alignment about the path we're on and why." From his perspective, it's that alignment, not technology, that's the biggest blocker. "It has to start with the why and work backwards. It's not about the what, it's about the why. There has to be alignment across the organization. It's the shift from handling just interactions to really owning the end-to-end member experience together."