SULSUL
AI-powered Customized Interview Preparation Platform
Project Overview
SULSUL is an all-in-one interview preparation service that allows users to practice interviews with customized questions generated based on their self-introduction. It supports systematic interview preparation through AI feedback, past question challenges, and live interview simulations.
Project Details
Development Period
2024.08 ~ 2024.11
commons.fronend
Key Features
AI Interview Practice
Resume-based Question Generation & Feedback
My Role
Front-end
UI Implementation, API Integration, Storybook Deployment
Project Screenshots
SULSUL Landing Page
Intuitive and modern design for the service introduction page
Interview Practice Page
Interface for practicing in an environment similar to a real interview
Question Selection Modal
Selection of customized interview questions generated based on self-introduction
Feedback System
AI-based answer analysis and improvement suggestions
Technical Challenges and Solutions
API Speed Issue Due to Free Server Migration (12s → 5s)
Problem
After migrating from a personal server to Koyeb, components rendering data received from API calls experienced a 12-second blank screen. Simple loading indicators did not provide a good UX.
Solution
Improved by using `usePrefetch` to pre-fetch data before rendering. Instead of loading all data at once, requests were made per self-introduction question unit to minimize blank screens. Auto-focus was implemented to immediately display pre-fetched questions, enhancing UX.
Minimizing Development Stages with Storybook Deployment for Design Verification
Problem
When sharing unverified work with designers, verification was only possible through a deployed site. This required deploying in an unstable state and going through deployment and verification stages just for design changes.
Solution
Improved by separately deploying components that required designer verification using Storybook and Chromatic, and receiving design feedback through this method. Design verification speed increased, development schedule coordination became smoother, and real-time feedback accelerated development.