I’m a backend engineer who enjoys building services with a focus on optimization and AI integration.
- Languages: Java, Python, SQL
- Frameworks: Spring Boot, Spring MVC, JPA
- Databases: MySQL, Redis
- DevOps: AWS (EC2, RDS, S3, ECS), Docker/Compose, GitHub Actions, Nginx
- Testing: Prometheus, Grafana, JUnit5, JMeter
- AI Tools: YOLOv5, GPT-3.5, koBERT
- Email: yndbsk9372@gmail.com
- Blog: pyounani.tistory.com
- Portfolio: Portfolio
Backend & Infrastructure | 2025.03 – 2025.10
- Tech Stack: Python 3.11, FastAPI, Celery, Redis, MongoDB, AWS S3, Docker
- Key Contributions:
- Built an asynchronous AI inference pipeline using Celery + Redis to prevent API blocking caused by long-running AI inference tasks
- Improved timeout and polling overhead issues by introducing a Redis Pub/Sub + SSE-based real-time event communication architecture
- Optimized GPU memory usage and inference stability by designing an input-based dynamic model loading strategy for AI models
Backend & Infrastructure Lead | 2024.03 – 2025.02
- Tech Stack: Java 17, Spring Boot 3.2, JPA, MySQL 8.1, AWS (EC2, RDS, S3), Nginx, Docker, JUnit5
- Key Contributions:
- Identified HikariCP connection pool bottlenecks during load testing and improved TPS and latency through thread pool, connection pool, and timeout tuning
- Discovered excessive queries caused by JPA relationship design through AOP-based observability and resolved them by redesigning the ERD structure
- Improved email verification latency by converting synchronous email processing into an asynchronous architecture and optimized thread pool behavior using Mock SMTP-based load testing
- Designed a custom RetryPolicy to selectively retry only specific email delivery exceptions instead of relying on generic exception-based retries
- Solved DB-S3 consistency issues during transaction rollbacks by designing an event-driven compensation and batch cleanup architecture
- Implemented social login and JWT-based authentication using Spring Security and OAuth2
Operations & Backend | 2025.01 – 2025.03
- Tech Stack: Java 17, Spring Boot, JPA, MySQL, Redis, AWS ECS, GitHub Actions
- Key Contributions:
- Optimized a popular-post ranking API by analyzing execution plans (EXPLAIN) and redesigning the ranking query into a 2-step DB ranking structure
- Improved infrastructure management efficiency by migrating to a managed container environment using AWS ECS
Thanks for visiting! 😊