Core Technologies
Applied in personal projects: SwiftUI, Combine, Swift Concurrency
- Built iOS CI/CD at Coinone and a shared CI/CD foundation at Hecto, eliminating local build dependencies so development could continue during deployments and any team member could deploy
- At Coinone, introduced Tuist and modularization standards, reducing build time by 10% and standardizing team project structure
- Independently planned, developed, and released 3 iOS apps after leaving Hecto, extending services to the web and Toss mini-apps with AI agents
Work Experience
Mobile app development for card and telecom value-added services. Owned 9 apps total - 5 iOS Native (RxSwift, MVVM) and 4 Flutter (Clean Architecture) - with experience across native iOS, Flutter cross-platform development, and solo app ownership.
- Solo development of global security iOS app for URL security checks and in-message URL analysis, from initial design through store release
- Safari Extension for real-time in-browser URL checks; Share Extension for URL sharing and analysis from external apps
- URL threat detection via Nurilab SDK integration
- WebView-based home screen; UIKit native screens for Extension setup guidance and app settings
- Built as English-only global app
- Designed and applied UIScene migration (SceneDelegate, Info.plist) and supported major iOS releases
- Handled bug fixes, feature improvements, and production issues
- Developed and released 4 new Flutter apps in a team of 2 Android and 2 iOS developers (Clean Architecture, Riverpod, go_router)
- Jointly designed shared architecture, implemented core features, and integrated platform-specific functionality
- Developed and released Family Care, EcoLetter, and Weather Stone concurrently in about 1.5 months; shipped Deepfake Protection in about 3 weeks
- Designed and built a shared GitLab CI/CD foundation for Android/iOS/Flutter apps and standardized build and release processes
- Documented AI Agent Rules/Skills operating model and shared with the team to improve collaboration consistency
- Real-time price and order book features with Socket + Protobuf
- WidgetKit asset and coin widgets
- Integrated an app security SDK for jailbreak detection
- Modularized the project with Tuist and standardized project structure and the build environment (10% reduction in build time)
- Built and operated Jenkins-based iOS CI/CD
- Participated in team code reviews before PR merge, managing code quality (SwiftLint)
- Led Swift 4.2 → 5.0 migration
- Implemented ID capture and facial verification flows with the Facephi SDK to support KYC compliance under Korea's Act on Reporting and Using Specified Financial Transaction Information (approximately 2 months)
- Built capture guide UI to show positioning and status during ID and face capture
- Moved ID verification checks to the server to prevent bypassing or tampering with client-side decisions
- Integrated an app security SDK for jailbreak detection, applying the same security policy as the exchange app
- Masked the screen when the app entered the background to prevent personal information from appearing in the app switcher
- Introduced Tuist's concepts and key features in a team presentation and adopted it in the project after reaching team consensus
- Seoul Metro collaboration: accessibility-focused subway prototype app
- WebView-based iOS app development and maintenance for Hyundai Capital (freelance)
- Knowledge and hobby learning platform connecting users with instructors; released on Google Play and App Store
- Beacon-based attendance Android and iOS app development
Personal App Development & Releases
Analyzes selected photos on-device for signs of AI generation and applies labels or blur to suspected AI-generated images on Safari webpages.
- Integrated the Smaug INT4 Core ML model and implemented the analysis pipeline in Swift, from image preprocessing through three-stage inference and score calculation
- Serialized inference requests with a dedicated actor and managed model memory lifetimes by loading and unloading models at each stage
- Implemented async/await-based photo analysis, task cancellation when leaving the screen, error handling, and retry flows
- Shared filter settings between the app and Safari extension through App Groups and synchronized image labeling and blur modes
- Built adaptive layouts for iPad screen sizes and orientations, with Dynamic Type and VoiceOver accessibility support
- Compared Swift preprocessing results with the original processing approach using 20 PNG images and verified the full inference flow on a physical iPhone
- Planned, developed, and released a 60-second timed quiz game where players guess Korean words from AI pun/association illustrations (Jun 2026)
- Directly developed the native iOS app (Swift, UIKit, Combine, MVVM); used AI agents to extend it to the web and Apps in Toss
- Separated game logic, timer, and answer validation in Domain-layer GameEngine; Repository pattern with 117 puzzles in puzzles.json
- Mode selection (10/20/50/full), initial-consonant hints, tier/best-record tracking, and Universal Links deeplink result sharing
- AdMob banner/reward ads, Firebase Analytics, and AASA (Universal Links) integration
- Solo development and App Store release of SwiftUI native app for solo-household chore routines (May 2026)
- Local persistence with Core Data and Repository pattern; Rule Engine separates routine cycle/due-date calculation from UI logic
- Onboarding, home, stats, and settings flows; local notification scheduling with per-routine on/off and permission handling
- Firebase Analytics custom events, Crashlytics, and AdMob adaptive banner (init gate) integration
- SPM dependencies, UI tests, and in-repo design/marketing docs for consistent UI and copy
- Planned, developed, and released Joseon-themed card game app on iOS/Android (Apr 2026)
- Difficulty and stage progression, card-matching game logic, sound and UI flow implementation
- Directly developed the native iOS app with Swift/UIKit; used AI agents to develop the initial React Native version and the Android Kotlin app
- Solo ownership of platform-specific store releases, maintenance, and native UX improvements
- Planned, developed, and released a Flutter app for daily guides, calendars, and connections on iOS/Android (Jan 2026); extended it to a Toss mini-app
- Organized state management and REST API integration with Riverpod, feature-based modules, and the Repository pattern
- Integrated Firebase Auth sign-in, FCM push notifications, deep links, iOS WidgetKit, and Android home widgets
- Implemented Korean/English localization, region-based routing, AdMob and Toss ad monetization, and Firebase Analytics/Crashlytics
Developer Tools & Automation Projects
Used AI agents for implementation while leading design, validation, and deployment.
- Built a Mac Mini home-server monitoring and recovery tool with remote metrics, Docker and health-check management; distributed via Homebrew
- Separated the monitoring daemon (miniopsd) from the client app for remote access, sharing Core/Monitoring/API through multiple SPM packages
- Developed an open-source Swift library for Korean profanity filtering with SPM installation, README examples, and test cases
- Supports rule-based filtering by default, with an iOS 26+ option that combines rule-based checks with Foundation Models-assisted detection
- Developed 4 mini-apps: Today Guide for You, Picture Puzzle, Will AI Take My Job?, and Magic One-Liner
- Integrated the Apps in Toss SDK, ads, analytics, and public-data APIs, with result sharing and per-app deployment packages
- Automated Reddit RSS collection, Gemini script generation, MeloTTS narration, subtitles, video composition, and YouTube uploads
- Prevented topic reuse with SQLite; deployed to a Mac Mini with Docker and managed daily scheduling, logs, and failure states
- Designed a signal, risk-management, and order-execution pipeline with loss limits, exposure caps, and decision audit logs
- Built staged validation from fee- and slippage-aware backtesting to walk-forward analysis, paper trading, and small-scale live trading
Education
About Me
Built iOS CI/CD at Coinone and a shared CI/CD foundation at Hecto Innovation. At Coinone, introduced Tuist in a team presentation before adopting it in the project and modularizing the codebase. At Hecto, prepared AI Agent usage guidelines and proposed approaches to modularization and documentation to the team. After leaving Hecto Innovation, I independently planned and shipped personal apps such as Picture Puzzle, SingleRoutine, and Amhaeng-eosa from concept to store release. I focus on turning repeated inefficiencies into standard processes, and whenever possible, I shape them into structures the team can actually use.
Additional
- Military: Army Sergeant, Honorable Discharge (Jul 2010 – Apr 2012)
- Certification: Information Processing Engineer / HRD Korea (May 2015)
- Availability: Negotiable upon final offer
- Personal Website: www.burdock12.site
- GitHub: https://github.com/wwwshe
- LinkedIn: https://www.linkedin.com/in/wwwshe
- Resume: https://www.burdock12.site/resume