Zowie Policies (Knowledge Base)
My end-to-end redesign for next-generation
AI knowledge management.
Knowledge Base 2.0 transforms chaotic multi-source management into a streamlined platform with automated sync and AI validation.
As Lead Product Designer, I spearheaded the complete redesign of Zowie's AI Agent knowledge management system, consolidating fragmented sources into a unified architecture that reduced daily management time.
My Role & Contribution
Zowie's AI Agent knowledge management system overwhelmed users with four different content types (Snippets, Articles, Websites, API), forcing knowledge managers to spend 45 minutes daily navigating fragmented sources and manually syncing content across regions, resulting in 72% AI response accuracy and a poor 2.9/5 user satisfaction score.
Problem
Successfully consolidate multiple knowledge sources into a unified, intelligent system that enables knowledge managers to efficiently maintain accurate, region-specific AI Agent content while reducing operational overhead and improving AI response quality.
Success definition
I conducted user testing with clients using Figma prototypes, revealing that users found dynamic AI validation "impressive" compared to static guidelines, while confirming websites as their most critical content source requiring real-time synchronization across multiple regions with localized languages.
Research & Discovery
I worked closely with PM, developers, and QA to ensure real-time API optimization for synchronization, scalable multi-language architecture, performance benchmarks under 200ms, and comprehensive accessibility compliance while maintaining feature complexity behind progressive disclosure.
Technical Collaboration
The redesign achieved 65% reduction in daily management time (45 to 15 minutes), improved AI response accuracy from 72% to 95%, increased user satisfaction from 2.9/5 to 4.6/5, and reduced knowledge-related support tickets by 62%.
Impact & Results