In this interview from Google Cloud Next 2026, Chris Mina, chief technology and product officer of LivePerson, joins theCUBE's John Furrier and co-host Alison Kosik to discuss how enterprises are overcoming "AI paralysis" to move from proof of concept to production-ready agentic customer experiences. Mina explains that while brands have made real progress building AI-powered self-service flows, many stall when edge cases surface and governance teams pump the brakes — a pattern he calls AI paralysis. To break the deadlock, LivePerson recently launched Syntrix, a testing platform that uses synthetic customers and generative scenarios to simulate hundreds or even thousands of complex conversation flows, giving brands the data they need to satisfy security, legal and AI council stakeholders before going live. The conversation also explores the orchestration layer at the heart of LivePerson's platform, including a guardian agent that monitors 100% of conversations in real time and continuously evaluates whether to escalate, redirect or let an interaction run its course. Mina shares a financial services case study where remote branch workers — previously underutilized due to the rise of online banking — are now being trained and deployed as digital agents, making far better use of an existing workforce rather than simply reducing headcount. He also outlines LivePerson's deepening relationship with Google Cloud, including early adoption of Gemini Enterprise and a multi-year migration of the full LivePerson platform to GCP — a move that has reduced support tickets, streamlined deployment and made Gemini an out-of-box model option for customers. From breaking AI paralysis with simulation-based assurance to repositioning branch workers as a digital-first workforce, Mina makes the case that CX remains the last great frontier for agentic AI, with enterprise direct-to-consumer adoption still in the single digits and enormous room to grow.