Knowledge Base Management for AI-Powered Customer Support.
A comprehensive knowledge management system that empowers support teams to create, organize, and maintain accurate content that powers AI responses and improves customer satisfaction.
As the designer leading this initiative, I owned strategy, UX, research, and cross-functional alignment for Zowie's next-generation AI knowledge management platform — taking it from a fragmented legacy system to a scalable foundation for future AI capabilities.
Problem
The knowledge management system I was tasked with fixing was split across multiple content types, forcing knowledge managers to manually maintain and sync information between regions. The result: operational overhead, inconsistent AI responses, and a satisfaction score of 2.9 out of 5.
Solution (What I did)
I redesigned the Knowledge Base end-to-end — consolidating multiple content sources into a unified, scalable architecture with automated synchronization and AI-powered validation. I introduced real-time updates, multi-language support, progressive disclosure for complex features, and a content management workflow that made the day-to-day feel significantly lighter.
Impact
The changes I designed delivered measurable, immediate results: Daily management time cut by 65%, AI response accuracy lifted from 72% to 95%, User satisfaction jumped from 2.9 to 4.6 out of 5, Knowledge-related support tickets dropped by 62%. Beyond the metrics, the new architecture I defined became the foundation Zowie's AI Agent capabilities are being built on.
