web design / proof of technology
Quiz Question Generator
An AI-generated quiz platform that automates question generation to reduce workload for medical faculty.
context
Project Overview
The Challenge
Faculty members want to create a low-stakes quiz platform for students for retention and understanding, but lack the time on top of all their other tasks.
The Solution
Designed the concept to production-ready prototypes of an AI platform that generates quiz questions with human approval.
my role
What I Owned
Workshop Facilitation
- →Led client workshop
- →As-is and To-be scenarios
- →Understanding course content (what the AI draws on)
Design & Prototyping
- →Wireframing + Engineer Hand-Off
- →Design Reviews
- →Style Guide Creation
Design Mentor
- →Directly mentored a summer intern through user flows, sketch work, prototypes, and client feedback sessions.
outcomes
The Quiz Platform
The solution was a dynamic, clickable prototype with an AI model implemented to help with question generation.
- Faculty can upload course material and learning objectives to base the generated questions off of
- Faculty can easily accept or reject questions to create a quiz
- Students now have the opportunity to quiz themselves on course material
- Students get immediate feedback on quiz performance
This solution follows department goals of a flipped classroom, where medical students gain initial exposure to new material outside of class and use class time for higher-order activities. Delivered as working code and design specs, the prototype became part of the client's grant submission.
Faculty Experience
Easily Manage Questions
When creating a quiz, faculty can provide example questions to guide the generation. They can also easily approve, deny, or edit generated questions.
See Class Insights
Faculty can view the insights tab, seeing areas of strength and needed improvement based on quiz results.
Student Experience
Flipped Classroom
Students can test their knowledge outside of class and spend class time reviewing concepts they need assistance in.
Course Insights
Students can see what concepts or categories they are strong in, and what areas they should work to improve on after taking the quiz.
methodology
The Process
Understanding the User
Facilitated a client workshop to identify current/desired experiences for end users (students, faculty), discussing platform goals, current pains, and success criteria. Also mapped out access to data sources (course content) that would be used to train the AI model.
Brand Guide + User Journey
To set the foundation for design, I built a Figma brand guide with variables for color, typography, and component variants for fast, consistent prototyping. I advocated for building within the client's design system, workshopped faculty and student user flows so the screens would reflect each group's real needs, and mentored an intern through the process.
Sketches to Prototypes
Turned sketches to prototypes for the faculty and student flow, using Figma Make for rapid prototyping, polish, and text generation. Led design reviews with engineers and clients throughout to align on feasibility and UI direction.
Engineer Hand Off
Coordinated with a front-end developer to turn the designs into production production-ready code. Worked with AI engineers to connect the front end to the back end (an AI model served through FastAPI endpoints that generated the quiz questions).