Cortex Signals: How AI Agents are Transforming Customer Experience (part 2)

Welcome to Cortex Signals, insights on AI from Cortex Reply. In this second episode, host Sanjay Dandeker continues his conversation with Matt Hull and Josh Prout, going deeper into the practical challenges and future possibilities of AI in customer experience and contact centres. The panel tackles how organisations should be testing AI systems using real interaction data rather than limited pre-written scenarios, and why single-number metrics like average handle time and CSAT scores no longer tell the full story. They explore deep observability, understanding not just what an AI said, but why it said it, and the growing importance of this in regulated industries like banking. The conversation also covers the underestimated complexity of knowledge management, with real-world examples of organisations discovering they needed thousands more knowledge articles than anticipated. Matt and Josh discuss how human agent roles are being elevated rather than replaced, and look ahead to capabilities such as proactive, personalised outreach, AI concierge services, and seamless auto-channel switching between voice and messaging platforms. Key topics covered: • Building AI test cases from real customer interaction data • Observability and deep interaction analysis in AI systems • Regulatory compliance and AI decision explainability • Knowledge base curation and management for AI • The evolving role of human agents as subject matter experts • Proactive outreach, concierge AI, and channel switching • The future of AI-to-AI interactions Subscribe to Cortex Signals for more insights on AI from Cortex Reply.