Minjae Kang.
Frontend Developer
Introduce.
I took a leave of absence from university to found a startup and began coding to build the service myself. Since then, I have planned RiderTown, a motorcycle maintenance and riding platform, and JustTrip, a collaborative travel platform, and built, launched and operated them as web and app products with React, Next.js and React Native.
Drawing on full-stack experience building everything from the frontend to the backend and infrastructure myself, I understand how data flows through the whole system, and I will contribute to the growth of the team and product through in-depth communication and relentless execution.
🏝️Rescue the developer trapped on the island writing his resume.
Rescue DeveloperExperience.
Built undoable optimistic deletion as a shared structure across 7 domains and cut refetches during consecutive deletes to one
Cut the trip detail screen's initial load in half (890kB → 452kB, 49%↓) and removed re-renders on map interaction
Solved out-of-order search results and duplicate requests under concurrent load, choosing the right approach for each case
Designed separate routing for Korea and abroad and reduced Directions API calls by merging consecutive same-mode legs into waypoints
Solved out-of-memory errors in bulk photo ZIP downloads by splitting them into 50-photo chunks
Built an offline mode so itineraries and vouchers of any trip opened once can be viewed in airplane mode or areas without a network
Designed access control with the backend so only trip members can view photos and vouchers, and applied it on the client
Designed Document AI-based auto-fill for vouchers and receipts with the backend and built automatic single/bulk form switching
Led the entire frontend process, from feature, spec and design system design to launching the Expo app and the responsive Next.js web
Set up an AI-agent development workflow with CLAUDE.md and DESIGN.md documentation and a remote MCP server built from the backend Swagger spec
Built a non-ferrous metal order/quote platform and admin dashboard, unifying quote and order channels scattered across phone and messengers
Built a real-time riding architecture that keeps going in the background by replacing WebSockets with Realtime Database and JS location collection with native background collection, with onDisconnect letting the server clean up presence, to fit riders who keep a navigation app open
Improved ride record accuracy and cut route transfer by over 70% with time-difference-based ride time, dynamic send intervals by speed and stop state with distance sampling and polyline encoding, and lean angle filtering at low speed (above 15km/h, 64° cap)
Migrated AWS MSA (EC2·RDS·VPC) to fully serverless Firebase to run the service alone after the backend engineer left, cutting fixed monthly infrastructure costs by over 90%
Pivoted from maintenance booking to real-time riding based on per-page visit metrics
Migrated a React SPA to Next.js, shrinking the initial bundle from 2.5MB to 550KB (~78%↓) and the WebView first-screen load time from 3.5s to 0.8s (~77%↓)
Built a Mediasoup v3 SFU server to fix group voice call uploads at one stream per device (3 → 1, about 67%↓ for 4 people)
Developed real-time riding records with gyroscope-based lean angle measurement and 1-second WebSocket crew record sync
Co-designed a Redis lock that atomically holds a slot for 5 minutes during payment, fixing double bookings of maintenance slots at peak hours
Validated the market with an on-site survey of 100 riders and personally built the web, app, Spring Boot backend, AWS infrastructure and Docker deployment automation to launch V1
AI Experience.
Document AI & Data Pipeline
2026.08 - 2026.09Intelligent Voucher & Receipt Auto-Parsing
Data Pipeline Engine Built on Azure Container App & Azure Document Intelligence
Simplified input (1 field for flights and lodging, 3 select forms for expenses) Reached 95% voucher parsing accuracy and cut registration time by 80% Reduced manual expense entry by over 90%
Computer Vision & Synthetic Dataset
2026.04 - 2026.05Synthetic Dataset Pipeline & YOLOv8 Object Detection Training
50,000-Image Synthetic Dataset Generation & Multi-Region Object Detection Fine-Tuning
Built a model that precisely detects in-game objects on live screens without manual labeling
LLM & Conversational Agents
2026.09 - PresentInteractive Resume AI Assistant
Real-Time Conversational Agent Powered by RAG Knowledge Injection & Prompt Engineering
Delivered a conversational resume that instantly answers with tech rationale and problem-solving evidence
LLM & Conversational Agents
2026.03 - 2026.04Vita - Personalized AI Supplement Recommendation Engine
Health Survey Structuring & Real-Time Nutrient Analysis via LLM
Launched a web and app service combining personalized supplement analysis reports with intake calendar routines
LLM & Conversational Agents
2025.09 - 2025.10HappyWhale - Emotion-Adaptive Daily AI Notification
Real-Time Empathetic & Motivational Messages Tailored to User Emotional States
Delivered personalized LLM emotional messages with high uptime, providing positive emotional care experiences to over 1,300 active users.
Award.
Korea SMEs and Startups Agency (KOSME).
2024.03 - 2024.11Youth Startup Academy (Startup Success Package).
Motorcycle Consumables Lifecycle & Maintenance Platform via Riding DB [100M KRW Scope]
Korea Institute of Startup & Entrepreneurship Development (KISED).
2023.05 - 2023.12Pre-Startup Package.
Motorcycle Repair Quotation & Community Platform [40M KRW Scope]
Other Experience.
RiderTown.
Co-founder / Field Sales
2022.11 - 2024.03Repair Shop Partnerships & Early User Acquisition.
2023.06 - 2023.08With no initial capital or marketing budget, we needed to secure repair shop partners and rider users at the same time.
Visited repair shops nationwide with business proposals to win partnerships Handed out QR business cards and introduced the service to riders at Jamsu Bridge and rider cafés Set up marketing partnerships with riding cafés, trading poster placement for café feature posts.
Onboarded 41 high-performance repair shops nationwide and acquired early rider users. Secured the Pre-Startup Package and Youth Startup Academy government programs.
Skills.
Overall.
Prefer clean user interfaces, polished visual hierarchy, and fluid micro-animations.
Follow new technology closely, but choose what fits the situation.
Actively explore unfamiliar areas when the work calls for it and deliver results.
Communication.
Proactively voice ideas, architectural proposals, and product improvements.
When encountering ambiguities or blockers, pinpoint the exact unknowns and formulate precise, constructive questions.
Always prioritize active listening and empathy before asserting my own viewpoint.
JavaScript / TypeScript.
Proficient in JavaScript and TypeScript.
Understand ES Modules-based tree shaking and cut unnecessary dependencies to reduce bundle size.
Define data contracts between components and modules with types to improve maintainability.
React / Next.js.
Extract repeated logic into custom hooks and repeated UI into shared components for reuse.
Proficient in designing optimal rendering strategies (SSR, SSG, ISR) and SEO structures in Next.js.
Manage global state with Redux, and server state and data caching with React Query.
Experienced in performance optimization through Core Web Vitals profiling, dynamic imports, and route-based code splitting.
Experienced with Jest unit tests and React Testing Library component tests.
React Native / Expo.
Capable of delivering iOS and Android apps from concept to store release with React Native and Expo.
Experienced in integrating native capabilities such as real-time GPS tracking, camera, and push notifications.
Skilled in automated build/deploy pipelines using CodePush and EAS, with rich experience navigating store reviews.
Java / Spring.
Able to design and build object-oriented RESTful API servers based on Java and Spring Boot.
Experienced in data modeling, transaction management, and query optimization based on Spring Data JPA and MySQL.
Applied distributed locking with Redis for concurrency control, such as booking slot reservations.
Experienced in event-driven asynchronous messaging using RabbitMQ.
AI / LLM.
Built content generation and AI features by integrating OpenAI and Gemini APIs.
Built MCP servers and designed AI agent rules with Skills.md and Rule.md.
Ran a vision model project end to end with YOLOv8 and PyTorch, from dataset construction to training.
DevOps.
Experienced in building and operating infrastructure on AWS VPC, EC2, RDS, S3, and CloudFront.
Experienced in containerizing development and deployment environments with Docker.
Experienced in automating builds and deployments with GitHub Actions.
Education.
Kwangwoon University.
2018.03 - 2026.08B.S. in Electronic Engineering
Sanbon High School.
2013.03 - 2016.02Natural Sciences Track