RDU · Durham
CFCI AI Intake System
Designed an LLM-powered intake system that intelligently matches startup requests with student consultants.
- Role
- Designer
- Timeline
- OCT 2025 – NOV 2025
- Technical Dialect
- Figma
- -67%Projected manual review time
- +21%Projected completed proposals
Problem
CFCI's project matching process was severely bottlenecked by a fragmented intake system reliant on outdated forms and unstructured emails, which frequently resulted in abandoned submissions and a low volume of viable projects. Even when applications were completed, critical technical details were consistently missing, making it difficult for the center to accurately pair projects with student teams.
Solution
I designed and built an AI-powered intake application that turns casual founder interviews into clean, standardized project briefs. By engineering an interactive extraction dialogue, the platform guarantees applicants provide all necessary technical and operational constraints upfront — eliminating dropped communication and equipping CFCI with the precise, complete data required to match projects immediately.
§ Shipped
What I built
The Process
Design architecture
Optional deep dive — how the problem became the shipped product.
User Research & Funnel Auditing
Interviewed CFCI staff and startups who had previously gone through the matching cycle to pinpoint exactly why applications were dropped and where critical engineering data was being lost in the old form-and-email workflow.
Design & Deployment
Executed the end-to-end product design and prototyped technical execution—mapping data-driven user flows in Figma and bridging them into production.