Voice Isn't Dead—It's Becoming the Foundation of AI Customer Experience

For years, enterprise voice infrastructure was treated like electricity - essential, but largely invisible. As long as calls were connected, few executives thought much about the underlying network.

That mindset no longer works. As AI reshapes customer engagement, voice is undergoing a quiet but important transformation.

The rapid adoption of AI agents, increasingly complex customer engagement platforms, and globally distributed contact center operations are transforming voice from a utility into a strategic application. AI is only as effective as the infrastructure supporting every interaction.

Voice infrastructure is now a strategic layer in modern customer experience (CX) stacks.      Barbara Dondiego, CEO of AVOXI, shares how cloud, AI, and evolving customer expectations are redefining the role of voice. She also explains how AI-ready cloud voice infrastructure is proving a competitive differentiator, and why IT and CX leaders are paying much closer attention.

Q: For years, telecom was viewed as a utility. Why is voice infrastructure becoming strategically important again?

Barbara Dondiego: Three major trends are changing everything.

The first is cloud transformation. We're taking something that was traditionally physical and moving it into software. Once voice becomes an application rather than just a phone line, organizations can attach data to every interaction, intelligently route calls, gain visibility into global traffic patterns, and optimize customer journeys in ways that simply weren't possible before. We're also seeing organizations consolidate multiple regional voice providers into a single cloud-based approach, giving them greater visibility, consistency, and operational control.

The second driver is AI. Organizations are adopting AI to lower costs, improve customer service, and increase operational efficiency.

Customers increasingly want to speak rather than type or text. In many cases, talking is simply faster and more effective. As AI enables more natural voice interactions, voice is experiencing a resurgence.

We’re seeing a rise in what I call “spoken service.” Customers expect to speak naturally and be spoken to naturally in return as well - whether they're interacting with a human agent or an AI agent - and receive immediate resolution.

The third trend is the growing need for secure, proactive infrastructure. As voice becomes more tightly integrated with AI, contact center platforms, and enterprise applications, organizations need greater visibility into performance and the ability to identify and resolve issues before they affect the customer experience.

Voice is the fastest path to solving complex problems. That's creating a resurgence in spoken customer service and making the underlying infrastructure more strategic than ever.

Q: Many organizations are focused on deploying AI. Why should they be thinking about voice infrastructure at the same time?

Barbara Dondiego: AI introduces tremendous complexity into every interaction.

A single customer call may now travel through multiple databases, AI services, analytics platforms, CRMs, and workflow engines before the customer even receives an answer. That means the quality of the underlying infrastructure becomes critical.

If there’s latency, delay, or poor audio quality, the entire interaction can fail. 

Today, human agents often identify these problems because they notice patterns during conversations and report them internally. Increasingly, those interactions may be handled entirely by AI. Going forward, organizations need data-driven visibility into performance so they can detect and fix problems proactively. Instead of relying on people to detect quality issues, enterprises need infrastructure that continuously monitors performance and identifies problems proactively through data.

Q: As AI agents become part of the workforce, what role does voice infrastructure play?

Barbara Dondiego: The customer is still making a phone call. That call may ultimately be answered by a human, an AI agent, or a human working alongside AI. Whether a call is handled by a human, an AI agent, or a combination of both, the experience still depends on the infrastructure. If the customer can’t hear clearly or there’s too much delay, the interaction breaks down.

That's why infrastructure matters. Voice quality, latency, packet loss, silence detection, routing intelligence, and other network characteristics all directly influence the customer experience.

There's another element that's often overlooked: phone numbers themselves. Customers associate local, recognizable numbers with trust. If an outbound AI contacts a customer from one number, but a follow-up human call arrives from a completely different number, that inconsistency creates confusion. Customers are more likely to answer and trust calls from recognizable, local numbers. Maintaining consistency across interactions—especially when AI is involved—helps build that trust.

Q: Everyone is talking about AI models. Are organizations paying enough attention to the infrastructure beneath them?

Barbara Dondiego: They need both. An AI model that doesn't understand customer intent obviously won't succeed. 

But even the best AI can't overcome poor voice quality. If the customer can't hear clearly, experiences noticeable delay, or doesn't trust the number that's calling them, the interaction has already been compromised.

You can’t separate the two. Successful customer experiences require both intelligent AI and high-quality voice infrastructure working together. This becomes especially important for large multinational organizations operating across dozens of countries with multiple contact center platforms, carriers, and regional phone numbering systems. That's where infrastructure orchestration becomes essential.

Q: Contact center technology stacks are becoming increasingly complex. What challenges does this create for IT and CX leaders?

Barbara Dondiego: We’re seeing a shift toward product-oriented roles in IT and CX organizations, where leaders are responsible for assembling a portfolio of tools, integrating them, and ensuring they work together seamlessly. That’s a much more complex challenge than the traditional single-vendor model.

The people responsible for contact center technology are increasingly becoming product managers rather than infrastructure managers. Instead of owning a single platform, they're managing portfolios of technologies—contact center platforms, CRM systems, AI applications, analytics, workflow engines, and communications infrastructure.

Their job isn't simply selecting technology anymore. It's making all those technologies work together.

AI has accelerated this trend because there are now dozens of specialized AI vendors solving different problems. Organizations are stitching together increasingly sophisticated technology ecosystems, and someone must ensure everything integrates, performs reliably, and delivers measurable business outcomes. That's becoming one of the biggest operational challenges facing enterprise CX teams.

Q: We hear a lot about AI orchestration. Does voice infrastructure become part of that orchestration layer?

Barbara Dondiego: I think about it a little differently. 

Voice infrastructure becomes the stable foundation underneath a rapidly changing AI environment. AI applications will continue evolving quickly; new models, vendors, and capabilities are expected.

Voice infrastructure is different. It’s complex to easily manage phone numbering plans, carrier relationships, routing policies, and global telecommunications standards if you are doing it manually and with analog processes. Once you've digitized and centralized that infrastructure, it becomes a dependable layer that everything else can connect to.

At AVOXI, we talk about orchestrating voice infrastructure, not orchestrating the entire customer experience. Our goal is to provide a centralized platform with the right control, visibility, and flexibility that customers can integrate with whatever AI technologies they choose today or adopt tomorrow.

Q: Some enterprises are separating voice infrastructure from their contact center platform. What advantages does that provide?

Barbara Dondiego: Primarily flexibility and control. When voice infrastructure is independent, organizations can change contact center platforms, adopt new AI solutions, or introduce additional technologies without rebuilding the entire communications foundation.

It also gives them greater visibility into infrastructure performance and allows them to optimize routing, quality, and global operations independently.

Perhaps most importantly, enterprises gain choice. They're no longer forced to access AI capabilities only through a contact center platform. They can integrate directly with AI applications while maintaining a consistent voice environment.

Q: Looking ahead three to five years, what will the enterprise voice stack look like?

Barbara Dondiego: We'll see much more cloud adoption. Voice infrastructure globally is still earlier in its cloud journey than many people realize. There's significant migration still ahead.

We'll also see far less hardware. Gateways, switches, and many traditional telecom components will gradually disappear as communications become entirely software-based.

The biggest shift, however, will be proactive infrastructure intelligence. Instead of waiting for customers or agents to report quality problems, platforms will continuously detect anomalies, identify issues at the phone-number level, and resolve many problems automatically before they affect customer interactions.

Beyond core infrastructure, capabilities like branded caller ID, security, fraud detection, synthetic call detection, analytics, and automated testing will become much more tightly integrated into the voice platform itself rather than existing as separate point solutions.

Q: What's your advice for organizations rethinking their voice strategy?

Barbara Dondiego: I'd encourage leaders to rethink voice in two important ways.

First, stop viewing voice as a utility. Think of it as a software application that can be orchestrated, optimized, integrated with AI, and used strategically to improve customer experience.

Second, stop evaluating communication channels solely based on cost. For years, the industry focused on deflecting voice calls because they were considered expensive. That's appropriate for simple requests. Customers should absolutely be able to check an account balance through self-service or receive a quick text message.

But when customers have complex issues, speaking with a person - or an AI capable of having a natural conversation - is often the fastest and most satisfying path to resolution.

The goal isn't to eliminate voice. The goal is to use the right channel for the right customer at the right moment. As AI reshapes customer engagement, that philosophy may become one of the most important competitive advantages enterprises can develop.

Voice is not disappearing - it’s evolving. Organizations that recognize its strategic value and invest in the underlying infrastructure will be better positioned to deliver seamless, high-quality customer experiences in an increasingly complex digital landscape.