Executive Principal Product Manager
Agentic AI Customer Service Platform
A top-three U.S. residential lender and fintech-driven company envisioned Agentic AI solutions that would consolidate fragmented support processes, reduce training time, and unlock millions in operational savings.
I owned the experience vision, design system, and end-to-end product design from early discovery and concept through detailed interaction design, prototyping, and pilot delivery,
Challenge:
For customer support agents, no two calls followed the same path. The design challenge was creating an interface that could guide them fluidly through a constantly shifting experience, so they could focus on having a genuine human conversation with the customer.
Solution:
AgentIQ was built on an agentic AI framework that actively orchestrated workflows in real time. As a call unfolded, the platform dynamically assembled the right information, triggered the right actions, and presented the right next steps based on the live context of the conversation
Challenge:
Support agents are on live calls with borrowers who may be stressed or frustrated. The design challenge was introducing AI-generated guidance into that high-pressure environment in a way agents would actually rely on. Recommendations needed to feel credible enough to act on quickly, but transparent enough that agents learned to trust the information that was surfaced.
Solution:
AgentIQ surfaced AI recommendations with clear context about where information was coming from and why a particular next step was being suggested. Rather than presenting guidance as a black box, the interface gave agents enough visibility to build confidence in the tool over time, and enough control to override or adjust when their judgment called for it.
Challenge:
After every call, agents were expected to manually document the interaction before moving on to the next one. Under constant pressure to keep queues moving, notes were often rushed, inconsistently structured, and filled with personal shorthand and errors. The result was a call record that was difficult to act on and nearly impossible to analyze at scale.
Solution:
AgentIQ automatically generated a call summary from the live transcript, producing an accurate, standardized record in the time it took an agent to glance at their screen and confirm it. Agents went from authoring notes under pressure to simply verifying an output that was already correct. In addition, standardized call records unlocked new capabilities for quality analysis and operational insights that had previously been out of reach.