AI in CX: Separating Hype from Impact: Sip & Swirl Roundtable Highlights

On Thursday, April 2, Liveops brought together a select group of CX and customer contact leaders for a virtual Sip & Swirl roundtable focused on one of the biggest questions facing the industry right now: how do organizations separate AI hype from real operational impact? 

Designed as an executive-only conversation, the event paired peer-driven discussion and fresh CMP Research insights with a guided wine tasting experience. Liveops sent each attendee a curated tasting kit ahead of the session, creating a more engaging setting for an otherwise serious conversation about AI investment priorities, operational readiness, and customer experience strategy. 

Guided by Nicole Kyle, Managing Director and Co-Founder of CMP Research, the roundtable explored what meaningful AI progression actually looks like across customer contact environments, from automation and generative AI to emerging agentic systems.  

Instead of focusing on buzzwords, the discussion stayed grounded in what leaders are facing now: pressure to modernize, growing expectations to prove ROI, and increasing urgency to make smarter decisions about where AI belongs and where it does not. 

AI investment is rising, but so is the pressure to get it right 

A major theme throughout the discussion was the urgency around AI investment in CX. Customer contact leaders are not just being asked to explore AI. They are being asked to operationalize it in ways that improve outcomes, reduce friction, and justify spend. 

Nicole Kyle framed the moment clearly, noting that the headline for CX strategy over the next two years is essentially “more self-serve and more AI tools.” But the conversation also made clear that the market is crowded, the terminology is noisy, and not every investment path will deliver meaningful value. 

One of the strongest findings from CMP’s research was the growing importance of customer analytics and insights. That stood out because it reflects a larger shift in how organizations are thinking about AI. It’s not just about adding tools. It’s about understanding customer behavior more deeply, improving journeys, and making automation work in a more purposeful way. 

As Nicole explained, better customer analytics, stronger knowledge management, and improved self-service are all connected. Organizations want automation to do more, but that only happens when the underlying data, workflows, and visibility are strong enough to support it. 

The challenge isn’t just the technology. It’s the people around it 

One of the most important parts of the discussion centered on a truth that often gets overlooked: even strong AI tools create new complexity. 

Nicole made this point directly, saying, “There’s a real acknowledgment that there’s a people consideration to our AI investments as well.” 

That insight resonated across the conversation. Leaders spoke candidly about what change management looks like in practice. For some, the challenge is organizational caution or workforce resistance. For others, it’s change fatigue, fragmented systems, or the difficulty of layering new capabilities onto already complicated environments. 

The conversation repeatedly came back to one idea: AI cannot simply be added into a customer operation and expected to succeed. It has to be introduced with trust, clarity, and usability in mind. 

Why agent assist is getting so much attention 

According to CMP’s research, one of the most active areas of AI investment is real-time agent assist. That finding sparked valuable discussion, particularly around why agent assist is emerging as a more immediate priority than more advanced autonomous systems in many organizations. 

The answer was practical. Agent assist gives organizations a way to improve speed, consistency, and agent support without fully removing the human from the interaction. It offers a more measured path forward, especially for leaders who want progress without unnecessary risk. 

Nicole described agent assist as an important response to the increasing complexity of frontline work. As simpler contacts move into self-service channels, live agents are increasingly left to handle more nuanced and higher-stakes issues. Agent assist helps close that gap by giving agents faster access to answers, recommendations, and next-best actions while still preserving human judgment. 

That distinction between augmentation and autonomy became one of the clearest through lines of the session. 

Agentic AI is generating interest, but not without caution 

The roundtable also explored the rise of agentic AI and where leaders are most open to enabling more autonomous capabilities. While interest is growing, the conversation was far from overly optimistic. 

Participants raised important concerns around hallucinations, governance, infrastructure readiness, privacy, and customer risk. In sensitive environments especially, the margin for error remains too small for some leaders to move aggressively into autonomy without strong safeguards in place. 

Nicole pointed to voice, chat, and assist tools as the areas where organizations currently show the most openness to agentic capability. Even so, the discussion suggested that openness does not mean haste. Leaders are weighing where autonomy can create real efficiency and where it could create new operational or customer experience problems. 

Customers still want one thing above all else: speed 

One of the most memorable insights from the session centered on what customers actually value most in service interactions. 

Nicole shared a finding that challenged a common assumption: “In live experiences, customers want speed.” 

That sparked immediate discussion because it cut through the long-standing belief that empathy is always the top differentiator. Instead, the group surfaced a more nuanced view.  

Empathy still matters, but not in the scripted, repetitive way many service experiences deliver it. 

Customers do not want empty apologies repeated back to them. They want agents and systems that understand the issue, move quickly, and make progress toward a solution. 

In that sense, speed is not the opposite of empathy. In many cases, speed is part of what makes an experience feel respectful. 

The metrics that matter are changing 

Another critical theme was measurement. As AI reshapes service delivery, leaders are also being forced to rethink which metrics still matter, which ones need to evolve, and how to tell a more complete performance story. 

Nicole noted, “What gets measured gets managed,” and that idea remained central throughout the conversation. 

CMP’s research highlighted customer satisfaction, self-service resolution rate, and first contact resolution as key metrics tied to customer loyalty. That’s notable because it reflects a shift away from relying too heavily on legacy measures alone. Leaders are increasingly being asked to evaluate AI not just by whether it reduces effort internally, but by whether it improves outcomes that customers can actually feel. 

Several participants also discussed the importance of separating live-service metrics from automated-service metrics. As more simple contacts move into self-service, live agents are left with more complex interactions, which changes how performance should be measured. 

Final thoughts 

Our Sip & Swirl roundtable was designed to spark a more honest conversation about AI in CX, and it did exactly that. 

What stood out most was not just the excitement around AI’s potential, but the maturity of the discussion around what it takes to make that potential real. The leaders in the room were not chasing trends. They were focused on practical application, organizational readiness, and outcomes that matter. 

Nicole captured the moment well when she pointed to one of the most difficult challenges organizations now face: “Managing change of an AI augmented workforce.” 

That challenge is real, and it may be one of the defining operational questions for CX leaders over the next two years. The organizations that get this right will likely be the ones that move beyond surface-level adoption and focus instead on trust, measurement, readiness, and customer impact. 

At Liveops, we believe those are exactly the conversations the industry needs more of as organizations work to turn AI ambition into meaningful, measurable progress.