Enterprise Connect 2026 brought together enterprise leaders across CX, architecture, AI, and transformation to unpack what’s holding organizations back from turning AI and CX ambition into real value outcomes. Alongside conversations on the expo floor, panels, and insight sessions, one theme kept coming up: AI in CX is moving fast — enterprise ways of working with it are not.
Here are five of the biggest takeaways from the event, through the lens of our team who were on the ground.
1. AI Strategy Is Still Being Driven by Emotion, Not Outcomes
Alexander Maximenko, Vice President of Solution Architecture, shares his take: “One theme that came up repeatedly at Enterprise Connect is that many organizations are deploying AI in the contact center using the same mindset they used for traditional software projects. They launch a bot, focus on containment rates, and assume the job is done. But if the bot can’t actually resolve the customer’s issue, all you’ve done is trap the customer in a faster IVR. At the same time, automating the easy interactions often leaves agents handling only the most complex and emotionally charged problems, while the AI itself is pulling from outdated or fragmented knowledge sources. AI in the contact center isn’t a ‘deploy and forget’ technology — it requires strong data foundations, connected systems, and continuous tuning to improve the customer and agent experience.”
2. Complexity Is Quietly Diluting CX Transformation
Adam Underkoffler, Vice President of Strategic Alliances, reflects: “One of the most consistent patterns we hear from enterprise teams is how often organizations end up buying overlapping CX and AI capabilities without realizing it. CCaaS platforms, CRM systems, cloud providers, and standalone AI tools are increasingly delivering similar functionality, and different teams often procure them independently. The result is duplicate investment, growing complexity, and fragmented customer experiences. In many ways, this reflects the broader convergence happening across the CX ecosystem, where multiple platforms are expanding into adjacent capabilities at the same time. The real challenge today is not choosing the wrong platform. It is navigating that convergence without creating unnecessary overlap.”3. Agentic AI Is Exposing the Limits of Legacy Delivery Models
Erik Delorey, Director of Innovation Engineering, comments: “A lot of organizations are trying to deploy LLM-driven and agentic AI using delivery models designed for traditional software. That mismatch showed up again and again in conversations at Enterprise Connect. AI systems don’t behave like static applications — they evolve, learn, and need continuous tuning. When enterprises apply rigid requirements, long procurement cycles, and “one-and-done” delivery thinking, they stand in the way of the very capabilities they’re trying to unlock.”
