Building backend systems with a focus on Architecture · Data · Reliability · Automation
From implementing features to understanding the architecture, data, and systems behind them.
I'm a software engineer based in Japan, primarily focused on backend engineering.
I enjoy understanding not only how to make software work, but also why systems are designed the way they are.
I prefer learning technologies by integrating them into real applications and improving those applications through small, reviewable changes.
| 📍 Based in | Japan |
| 🧭 Core Direction | Backend Engineering |
| 🔧 Current Focus | Architecture · Databases · Testing · Reliability |
| 🚀 Growing Toward | Cloud · Distributed Systems · Applied AI |
| Area | Focus |
|---|---|
| Backend Engineering | C# · .NET · Java · TypeScript · SQL |
| Deepening | Architecture · Database Design · Testing · CI/CD |
| Expanding Into | Cloud Infrastructure · Performance · Observability |
| Exploring | Distributed Systems · AI-enabled Software |
Evolving a simple ASP.NET MVC CRUD application into a structured and maintainable backend system.
This project is an incremental backend refactoring study focused on understanding why architecture and infrastructure components are introduced.
Architecture · Persistence · Dependency Injection
Testing · Error Handling · Maintainability
- Controller → Service → Repository architecture
- Repository abstraction and Unity Dependency Injection
- Entity Framework 6 + SQL Server LocalDB
- Code First Migrations
- DTO / Entity separation with AutoMapper
- Service-layer validation
- Global HTTP 400 / 404 / 500 exception handling
- Application logging with log4net
- MSTest service-layer unit tests
- Server-side pagination with validation and boundary handling
- Vue.js frontend integration
Current: Server-side pagination completed Next: Search / Filtering
Development trail:
➡️ Explore Dotnet Backend Study
Helping seniors and their families stay connected through asynchronous video communication.
An 8-member Java/JSP team project where I primarily contributed as a development / system design member.
- Designed application and server-side processing flows
- Structured the Servlet → Logic → DAO processing model
- Designed video playback and automatic reaction-recording behavior
- Defined unread / read state processing rules
- Created implementation specifications
- Designed functional, boundary, and exception test cases
- Reviewed implementation against agreed specifications
- Coordinated design and specification decisions within the team
The public repository also separates credentials and runtime-generated data from source control through environment-based configuration and repository hygiene practices.
Java 21 · Jakarta Servlet · JSP
JavaScript · MySQL 8 · Apache Tomcat 10
A social dining platform connecting people through shared meals and local events.
A 4-member team project built as separate frontend, backend, and AI-service repositories.
- Initial entity design
- Authentication flow
- Album backend
- Chat-room entity relationships
- Comment DTO fixes
- S3 image deletion behavior
- Next.js frontend and NestJS backend
- PostgreSQL + TypeORM persistence
- JWT and Google / Kakao OAuth
- Real-time chat with Socket.IO
- AWS S3 image handling
- FastAPI AI-service integration
- Docker and GitHub Actions
- AWS ECS / ECR deployment workflow
Next.js · React · TypeScript · NestJS
PostgreSQL · TypeORM · FastAPI
Docker · AWS · GitHub Actions
➡️ Frontend Repository ➡️ Backend Repository
Reconstructing a 2024 AI hackathon prototype as a privacy-safe and reproducible software project.
The current repository is an independent reconstruction using synthetic CRM data and does not contain the original corporate dataset or proprietary source files.
- Public / private data boundaries documented
- FastAPI health endpoint and automated test added
- Historical lead / note data relationships reconstructed
- Public aggregate CRM profile documented
- Privacy-safe synthetic data generator calibrated to observed aggregate behavior
- Leakage-safe machine-learning evaluation
- Reproducible preprocessing and feature pipelines
- Prediction and explanation APIs
- React / TypeScript dashboard
- Human-reviewed LLM-assisted outreach
- Automated tests, Docker, and CI
🚧 Current phase: calibrated synthetic data foundation ➡️ Next: leakage-safe preprocessing · reproducible EDA · model baselines
flowchart LR
A["Build Features"]
--> B["Design Boundaries"]
--> C["Improve Reliability"]
--> D["Scale with Cloud"]
--> E["Apply AI"]
style C fill:#d1fae5
My current focus is moving from simply implementing features toward understanding the larger engineering concerns around them.
Learning, working, and adapting across different countries and cultures.
| Country | Experience |
|---|---|
| 🇯🇵 Japan | Software engineering career · Japanese IT environment |
| 🇩🇪 Germany | 1-month intensive AI training · ML, Deep Learning, Azure OpenAI · Hackathon winning team |
| 🇺🇿 Uzbekistan | 1-week international internship · Cross-cultural professional experience |
| 🇷🇴 Romania | 1-month independent stay · Romanian language learning · Everyday life in a European environment |
| 🇰🇷 Korea | Korean background · Software engineering education · Team and personal development projects |
These experiences have strengthened my adaptability, cross-cultural communication, and ability to work in unfamiliar environments.
| Language | Level |
|---|---|
| 🇰🇷 Korean | Native |
| 🇯🇵 Japanese | C1 · Business level · JLPT N1 |
| 🇬🇧 English | B2 · Business level · TOEIC Speaking AL |
| 🇷🇴 Romanian | A1 · Beginner · Currently learning |
I prefer learning technologies by integrating them into real applications rather than studying them only in isolation.
When introducing a new technology or architectural pattern, I try to understand:
Why is it needed? What problem does it solve? Where should it belong? What trade-offs does it introduce?
I value incremental improvement:
Build → Find a limitation → Understand → Improve → Test → Review → Repeat
My goal is to grow into an engineer who understands both implementation details and the larger systems around them.