Last week’s “Genesys launches agentic orchestration features,”which focused on the product announcements and corporate messaging delivered during the keynote session atGenesys Xperience 2026. On the second day of the conference, Scott Cravatto, chief customer officer, Genesys, led a keynote built around Genesys customers telling their AI success stories live.
The customers provided a practical balance to Genesys’s ambitious agentic orchestration message. None of the customers started their AI journeys with autonomous AI everywhere. They began with choosing a problem, consolidating fragmented technology, connecting data and context, all while involving employees and risk teams.
Across the stories told, several themes emerged that go beyond technology, instead centering on some of the day-to-day challenges and implications each company faced. In addition to the customers presenting live at Xperience 2026, customer videos were sprinkled into the keynote on day one, and I included their stories in my review.
What problems can AI solve?
Several customers said the first challenge was deciding which problems AI should solve. Ferrari began with lead qualification, developing a model that achieved 96.7% precision in identifying low-potential leads. Technically, the model worked. But Rebecca Mequin, head of customer experience at Ferrari, said the company eventually recognized that model accuracy was not the business objective. Building better customer relationships was the end goal.
Ferrari expanded its thinking beyond a single lead or vehicle purchase to the customer’s entire relationship with the brand, including configuration, dealerships, events, service, merchandise and future purchases. “We didn’t need just to make one better decision about an incoming lead,” Mequin said. “We needed to make thousands of better decisions across the entire customer relationship.”
Best Buy Canada developed a three-question framework for evaluating potential AI projects: Is this the right work for AI? Does it make employees better? Can the company continue learning and improving from it? “If an initiative didn’t answer those three questions, it didn’t make the cut,” said Mick Robinson, director of customer experience, Best Buy Canada.
That framework resulted in Best Buy choosing high-volume activities carrying relatively little customer experience risk, such as inventory inquiries and routing, for their initial AI projects. Robinson reported that Best Buy’s natural-language routing and virtual agent deployments essentially eliminated misdirected interactions, while virtual agent chat reduced live agent traffic by more than 30%.
Computacenter offered an even more succinct test. “The goal isn’t to automate,” said Yomi Akinsanya, director, digital experience services. “The goal is to remove effort.” Its approach starts by using experience data to find friction, determining whether the problem can be prevented, then considering whether AI can augment employees. Automation is the final option rather than the starting point.
Consolidate technology and connect context
For several customers, scaling AI required first addressing fragmented platforms and disconnected customer data.
Before choosing Genesys Cloud, Best Buy Canada operated three platforms through three outsourcing partners, with little consolidated visibility. About 20 percent of calls were being sent to the wrong queue, and when contact volume doubled when stores temporarily closed during the pandemic, the scale of the problem multiplied. “We could not grow from that. We needed a scalable platform that we owned,” Robinson said.
Utility Warehouse described a similar experience with point AI products. “There’s often a hidden cost and challenge that people don’t talk about,” Kevin Mythen, product lead, Utility Warehouse saidhat data was available. “We can see where they (customers) are coming from, what they are doing, if they’re crossing channels – something that we were completely blind to when we had point solutions,” Mythen explained
The additional context supported a transition from bots that provide information to virtual agents that complete tasks. “We’ve gone from a very informational bot experience to an action model,” the Utility Warehouse representative said. Its virtual agents collect information, create tickets and perform work, producing containment rates of 30 to 50 percent, with some trials reaching 70 percent, Mythen said.
Jamie Timm, global senior vice president of service delivery and operations at TELUS Digital, cautionedthe experience worse. “The downside of poor orchestration,” she said, “looks like a heavier lift for the customer, starting over instead of that context being shared across the board.”
Bring employees, governance and scale together
During live sessions and on videos, Genesys customers consistently described AI deployment as an organizational change, not simply a technology implementation.
Auto Club Enterprises (ACE) moved 10,000 agents to Genesys Cloud in two months, but Jo Anne Tan, vice president, information systems, Auto Club Enterprises, said the company did not treat that milestone as the finish line. “The biggest shift was really around the change in the operating rhythm to introduce continuous improvement, learning and innovation,” Tan said.
ACE established weekly meetings with sales executives, monthly senior-leadership sessions, executive business reviews and AI showcases. Managers also became central to adoption, explaining why workflows were changing, listening for friction and translating employee feedback into improvements.
Tan emphasized that ACE applies human oversight according to the stakes of an interaction. Routine work such as summaries can be automated, but people continue to lead in emotional, complex, regulated or judgment-based situations. “The real decision is not AI or human,” Tan said. “It’s designing the model in such a way that you have the right partnership between AI and people.”
DNB similarly puts people at the center of its operating model. The Norwegian bank migrated 3,400 users from on-premises systems to the cloud over 12 months while activating Copilot, gamification and work automation. “Our operating model is 10% AI, 20% technology and the last 70% people,” said Niro Roman, product lead, DNB.
DNB also embedded compliance and risk specialists within the implementation team rather than treating them as final-stage approvers. Roman said an internal AI capability that conducts multiple risk assessments reduced work that previously took months to a few weeks. Overall delivery time fell 30%, while Copilot reduced after-call work by 50%.
Rabobank’s advice captured the broader message from the customer presenters: “Start small, think big, but scale fast,” said Thom Kokhuis, senior vice president, conversational AI Data & CRM, Rabobank. The bank now supports 20,000 employees with Genesys copilots and no longer summarizes interactions manually. Janelle Binder, senior vice president – marketing and customer success, Genesys, remarked that Rabobank has the largest number of employees using Genesys Copilot across the entire Genesys customer base.
But Kokhuis emphasized that scale depends on more than the platform: “It’s not only about technology readiness; it’s also about human readiness.”
An evolving orchestration platform
So, what are these Genesys customers using to get the results they were willing to publicly report at Xperience 2026? Genesys Cloud customers are already using Agentic Virtual Agent (AVA), and the integration ofScaled Cognition’s APT-2large action model (LAM), new development tools, Contextual Intelligence and the AI Control Plane are available. But many of the new capabilities Genesys talked about at Xperience 2026, including additional voice enhancements, A2A interoperability and AI-assisted authoring are scheduled for later, with Navigator and Orchestrator further out.
Customers are producing results with the technology that is shipping today, but defining components in the architecture that Genesys is hoping will redefine its category from CCaaS to an “agentic operating model” are not yet generally available.
In the world of AI, vendor pitches are no longer about selling “what’s on the truck.” What is important is offering existing and prospective customers a reason to believe that the Genesys product vision can help their company navigate a rapidly changing AI market.
