Backend-focused developer interested in AI agent systems, automation, and practical workflow engineering.

I like building systems that reduce repetitive work and make operational flows easier to reason about. Recently I have been focusing on LangGraph-style agent orchestration, QA automation, and backend architecture that can grow without becoming hard to maintain.

• Backend architecture with PHP, Laravel, Node.js, and Python
• AI agent workflow design with LangGraph / LangChain
• Browser, API, and DB-based QA automation
• Speech evaluation and AI feedback pipelines
• Practical automation for developer productivity

LangGraph-based multi-agent workflow engine inspired by JARVIS-style task orchestration.

• Backend: Python, FastAPI, LangGraph, prompt-based tool execution
• Frontend: Next 16 workflow visualization and execution trace viewer
• Monorepo: original integrated version

Android capstone project for sharing study records and code in an SNS-style mobile experience.

• Kotlin / Android
• Firebase
• MVVM, ViewModel, LiveData
• Feed, profile, comment, follow, notification flow

Older repositories documenting web fundamentals through practical implementations.

• JSP BBS: login, session, board CRUD, JDBC
• JSP Lecture Evaluation: JSP-based lecture evaluation UI/service flow
• Hackers Test: PHP account flow, Ajax, layout separation

Backend PHP, Laravel, Node.js, Python, FastAPI, REST API
Frontend React, Next.js, JavaScript, TypeScript
AI / LLM LangGraph, LangChain, GPT/Claude orchestration, RAG design
QA Playwright, Chromium automation, API/DB validation
Infra AWS Lambda, SSM Parameter Store, Docker
Database MySQL, PostgreSQL

• Prefer maintainable structure over trendy implementation.
• Keep orchestration state explicit.
• Separate controller, service, agent, graph, and tool boundaries.
• Design automation around actual operational bottlenecks.
• Document repositories so future readers can understand intent quickly.