Designing AI that Earns Trust and Relevance

Описание: What if doing “the right thing for this customer, right now” became the default? We sit down with Rob Walker, VP of Decisioning and Analytics at Pega, to unpack how empathy at scale, AI-driven decisioning, and strong governance can make customer engagement both more human and more effective. Rob explains why empathy isn’t charity; it’s a system for earning relevance and trust that compounds into mutual value. We dive into one-to-one decisioning that prioritizes a hierarchy of needs—solve hardship and service first, consider offers only when appropriate—and how real-time redecisioning keeps conversations useful as context shifts. The conversation gets candid on hype vs. reality: generative AI is expanding the content library needed for true personalization, even demonstrating better bedside manner in some settings. But trust requires more than sentiment; it needs policy. Rob breaks down the T‑Switch, a transparency and model-use framework that enforces explainability where stakes are high and speeds experimentation where stakes are low. Then comes the big horizon shift: customer advocacy agents. These AI agents will comparison-shop, negotiate, and handle service on our behalf, impervious to glossy ads and tuned to our preferences. That future could disintermediate brands and push competition toward price, reliability, and fit—raising the bar for authenticity and interoperability. Inside the enterprise, tools like Pega’s Customer Engagement Blueprint compress months of strategy design into hours, changing roles rather than eliminating them. Marketers and data scientists move from manual assembly to orchestration, governance, and measurement, while empathy and trust become operational standards, not slogans. If you’re ready to trade segments and batch campaigns for moment-level decisions and measurable trust, this conversation is your fast start. Subscribe, share with a teammate who needs to hear it, and tell us: what would you ask your customer advocacy agent to do first?