I'm a Product Engineer who turns complex policies and workflows into intuitive, scalable systems. I thrive on creating meaningful applications that solve real-world problems, participating in all stages of project development, from planning and design to implementation and deployment.
Role: UX Engineer Intern - Get more information about UX Engineer by Google UXE!
Period: 2025.03-2025.08 (6 months)
Ordinary Magic is a company making toys for infants and toddlers under the Peekaby brand.
Alongside its physical products, the company provides three software platforms: a mobile app, a subscription web service, and an internal admin console.
-
Mobile App
Provides play guidance and age-appropriate activity suggestions, helping caregivers understand how to use each toy and support meaningful play experiences with their children. -
Subscription Web
Allows caregivers to manage toy subscriptions, including delivery schedules, product information, and account-related settings. -
Admin Console
An internal platform used by operators to manage products, content, subscriptions, and user data, supporting daily operations across the entire service.
During my internship, I worked across multiple areas of the product, supporting both day-to-day operations and long-term platform improvements.
My work ranged from small but essential tasks to large-scale system design and refactoring efforts.
- Implemented and maintained UI components using the internal design system
- Improved usability and consistency across admin pages
- Fixed bugs related to permission handling and data visibility
- Refactored existing components for better readability and reusability
- Implemented form validation and safe default UX patterns
- Improved developer experience by cleaning up legacy code
- Participated in sprint planning and feature discussions
Among these contributions, I led two major initiatives that significantly improved both system scalability and operational efficiency.
As Peekaby’s platform expanded to Japan, the existing admin system began to show structural limitations.
- Permission logic was scattered across components
- Role-based visibility was difficult to reason about
- Adding new admin features required repetitive logic
- Risk of exposing sensitive data increased with new roles
- Glocalization Korean based system
I took ownership of redesigning and refactoring the admin console’s authentication, authorization, and overall architecture.
- Designed a unified admin role system
superAdmin,admin,external
- Implemented Firebase Authentication–based admin management
- Managed roles using
customClaims.role - Avoided storing admin accounts in Firestore
- Managed roles using
- Built centralized permission-checking utilities
- Refactored legacy admin pages into a consistent structure
- Implemented route-level and menu-level access control
- Standardized admin layout and navigation logic
- Introduced Smart / View component separation
- Reduced duplicated permission logic across the codebase
- Clear and predictable admin permission model
- Safer data access with minimized exposure risk
- Faster onboarding for new admin features
- Improved maintainability and readability of the admin codebase
- Stable foundation for future overseas expansion
Peekaby aimed to introduce a referral system to encourage organic user growth while preserving trust between caregivers.
This required careful design to ensure:
- Fair reward attribution
- Abuse prevention
- Clear user-facing UX
- Accurate tracking across multiple user states
I participated in designing and implementing the referral system from both product and technical perspectives, working closely with CPO to define rules and edge cases.
Before designing the referral system, I conducted UX research to understand how and why nurturers recommend products.
- Designed research goals and interview guides
- Conducted user interviews on real recommendation experiences
- Analyzed qualitative data to identify key referral triggers
- Compiled insights into a UX research report
- Applied findings to referral flow and system design
- Implemented referral code generation logic
- Designed referral relationship data structure
- Built referral tracking APIs
- Handled edge cases:
- duplicate sign-ups
- self-referrals
- delayed account completion
- Integrated referral logic with existing user flows
- Implemented admin-side monitoring tools
- Ensured data consistency between client and server
- Enabled measurable organic growth through user referrals
- Provided CTO with clear visibility into referral performance
- Prevented common referral abuse scenarios
- Delivered a scalable foundation for future reward campaigns
-
Frontend
React, TypeScript, React Native, Flutter
Design System Integration, Component Architecture, State Management -
Backend / Platform
Node.js, Spring Boot, RESTful APIs,
Firebase Authentication, Firestore, Cloud Functions,
Authentication & Authorization, Pagination & Data Modeling -
Database
MySQL, Firestore, Redis
Schema Design, Access Control Modeling, Data Consistency -
Testing & Quality
Vitest, Firestore Security Rules Testing,
Role-Based Permission Validation -
UX Engineering
UX Research (Interview, Qualitative Analysis),
UX Flow Design, Error Prevention UX,
Admin UX & Operational Tooling -
Tools & Collaboration
Git, Docker, Postman, Figma, Notion, Jira
Agile Development, Cross-functional Collaboration
I am open to new opportunities and collaborations. Feel free to reach out!
- Email: ihansaem1@kaist.ac.kr
- GitHub: GitHub Profile
Thank you for visiting my portfolio! 😊