Key takeaways:
-
Why production is the real test for AI in CX.
-
How to look beyond polished demos and assess whether an AI Agent can perform reliably in live customer environments.
-
Why Voice-first AI needs a different architecture.
-
Why latency, transcription quality, interruption handling, escalation and resilience make voice more complex than digital automation.
-
What “production-ready” really means.
-
The technical and operational requirements for deploying AI Agents safely, including orchestration, integrations, security, observability, reporting and service continuity.
-
How to adopt Voice-first AI safely.
-
How to start with targeted use cases, apply appropriate governance, measure outcomes and scale through a phased rollout.
Meet the Panelists:
ALEXANDER EATEDALI
Vice President and Head of UC and Core Services Engineering, Vonage
Alex leads engineering across mission-critical real-time communications and core services at Vonage. With deep experience in architecture, network infrastructure, security and operational resilience, he brings a practical view of what it takes to run voice and communications platforms at enterprise scale.
ALAN BECK
Chief Technology Officer, Syndeo
Alan has over 20 years’ experience designing and delivering voice, conversational AI and customer experience solutions for major brands including American Express, Voya, United States Postal Service, BSkyB and Hutchison 3G. With senior technical roles across Kainos, SpeechStorm, Genesys and Syndeo, he brings deep practical expertise in how to blend AI and voice technologies in real contact centre environments, at scale.
OLIVER LENNON (Webinar Host)
Co-Founder & CEO, Syndeo
As CEO at Syndeo, Oliver is responsible for driving the strategic direction of the company’s AI communications platform, which powers digital communications with AI and messaging. Syndeo specialize in Customer Experience consulting and technology, with a proven track record in delivering for world-class brands.
Webinar Overview:
Most AI for CX looks impressive in a controlled demo. The real test is whether it works in live voice environments, with real customers, legacy systems, compliance requirements and service teams depending on it.
This session goes beyond the AI demo to explore what it takes to build Voice-first AI Agents that perform in production. Voice AI is harder than digital automation: customers speak naturally, interrupt, change direction, get frustrated and expect fast answers. Behind the scenes, the platform must handle latency, transcription quality, security, integrations, escalation paths and operational resilience.
In 2026 and 2027, the winners will not be the organisations with the most ambitious AI demos. They will be the ones that can deploy governed AI Agents safely, measure outcomes and scale from targeted use cases into broader front door automation.
The discussion will bring together perspectives from Vonage and Syndeo, covering real-time communications infrastructure, AI orchestration, LLM guardrails and the practical steps needed to move from legacy voice self-service to production-ready AI Agents.
The goal is simple: help CX, operations and technology leaders separate AI theatre from AI that delivers measurable ROI in live customer operations.