CR CX Convo: What’s Next With AWS - A Sit Down with Amazon Connect’s Pasquale DeMaio

For anyone who follows Constellation analyst Liz Miller's research, the premise is not new. She has been talking about Amazon Connect as an engagement backbone for years, but her position has remained the same: it is more than a contact center platform. It is the infrastructure layer that enables enterprises to engage with the humans they serve, whether those humans are customers, employees, candidates, or patients.

When Liz sat down with Pasquale DeMaio, VP of Amazon Connect, for this CR CX Convo, the takeaway was confirmation that the floodgates have officially opened, and enterprises still thinking of Connect as a call center tool are already behind the curve.

Deflection Is Dead. Full Stop.

Pasquale said it better than most analysts have. If your CX strategy is still built around deflection, containment, and average handle times, you are not solving a customer problem. You are solving a cost accounting problem. In a world where a customer can open ChatGPT and ask it to solve their issue, the moment you deflect them, you have handed them off to someone who actually wants to help. How long before your business is just option C on a list?

The shift Pasquale is describing is from optimization to problem solving. Not how do we get customers off the phone faster, but how do we actually resolve what they came for? That sounds obvious. It is not obvious in practice. It requires a complete rethinking of what success looks like, from metrics to incentives to how you design the experience itself.

The People Problem Was Never the People

This is the thread Liz keeps pulling on in every conversation about enterprise AI, and it showed up here in a big way. For years, the dominant approach to CX technology has been to treat people as the bottleneck. Agents talk too long. Recruiters take too long. Doctors spend too much time on documentation. The hammer has been pointed at the humans.

What Amazon Connect is doing differently is designing around what people are actually good at. Not replacing them. Not deflecting around them. Getting the technology out of their way so they can do the thing they showed up to do.

The story Pasquale told about the heart rate monitor study is the one that stuck. They asked agents to wear monitors while testing new software designed to reduce the administrative burden of their jobs. Heart rates went down. Stress levels dropped. And a less stressed agent does what? They focus on the customer. They solve the problem. They do the job they actually want to do.

That is not a small thing. That is the entire design philosophy, and it is the right one.

Hiring Is One of the Best Use Cases You Are Not Talking About

The talent product deserves more attention than it is getting in the broader conversation about where agentic AI delivers enterprise value.

Think about what traditional recruiting looks like at scale. Eighty candidates to make one offer. Thirty days on average to close that process. Meanwhile, the best candidates are hearing back from five other companies, and the window to make an offer they will accept is closing fast. Layered on top of all of that is the reality that human beings carry preconceived notions into every decision. Accents. Zip codes. Whether the recruiter slept well the night before. All of it seeps into decisions that are supposed to be about whether someone can do the job.

What Connect's talent solution does is restructure that entirely. The assessment is about capability. The interview is about fit. The recruiter gets clean, structured data without the noise, and they can move. Amazon Connect is running this process in under a day. For any enterprise dealing with mass seasonal hiring, that is not an incremental improvement. That is a completely different model.

The Real Opportunity Is in the Connections

This is where CxOs need to pay attention, because it is a conversation that hasn't happened enough yet.

Each of these use cases, contact center, hiring, healthcare, and supply chain, operates as its own intelligent system. But the moment you start connecting them, the value compounds in ways that are hard to overstate. Supply chain decisioning tells you that Christmas demand is starting two weeks earlier than last year. That changes your seasonal hiring timeline. Hiring data flows into your HCM. Customer profile data unifies across every touchpoint, so that marketing, operations, and the contact center all learn from the same signal.

Pasquale made a point worth underscoring: this is the moment to bring the CMO into the Connect conversation. Not because Connect is a marketing tool, but because the data flowing through this ecosystem is exactly what marketing needs to be genuinely proactive. And proactive does not mean a chatbot popping up on a website. It means understanding what a customer needs before they have to ask.

The enterprises that figure out how to string these pieces of intelligence together will have a structural advantage that is very hard to replicate.

Questions You Should Ask Right Now

If there is one question worth walking away with, it is this: where in your organization are people still being treated as the bottleneck, and what would it look like to redesign that?

Not with AI for the sake of AI. Not with deflection dressed up as self-service. But with technology that actually gets out of the way so people, customers, candidates, and patients can do what they came to do.

That is the premise on which Amazon Connect has been building for years. The difference now is the scale and the specificity of what is possible. The floodgates are open. The question is whether your organization is ready to walk through them.