BACK
AI-Powered Knowledge Hub for Public Policy
An AI-powered knowledge hub for public policy that eliminates hallucinations through a RAG architecture delivering verified, cited answers from expert content in seconds.
My Role:UI/UX Designer
Platform:WordPress MVP · RAG
Timeline:4-Week Sprint
0%
RAG
MVP
3
01 — Challenge
A library full of expert knowledge that nobody could find.
The Centre for Civil Society had a growing library of expert content, but users couldn't find answers. They needed a modern solution that was fast, trustworthy, and resistant to AI "hallucinations."
The Core Problems:
Discoverability: Standard search failed on natural language queries.
Trust: Standard LLMs (like ChatGPT) invent answers; the client needed strictly verified answers.
Technical Gap: The dev team needed a clear path to build this on WordPress.
02 — Product Strategy
What gets built in 4 weeks is a product decision, not a design decision.
Before touching Figma, I ran a prioritization exercise mapping every candidate feature against user value and technical effort. The output determined the MVP scope ruthlessly.
The MVP Decision
The matrix made the call clear: User Accounts and Complex Filters consumed high dev effort for low immediate value. Removing them let us ship the core RAG experience in 4 weeks instead of 10.
More