Builds AI-powered products on top of backend and data engineering foundations.
Interested in the point where LLMs evolve from simple features into reliable production systems.

  🧩   Build    Backend + Data + LLM Products
  🌱   Explore    Reliable AI Systems · AI-assisted Developer Workflow

🥕 Karrot (Danggeun Market) · Data Engineer Intern · 2026.01 ~ 2026.04

Built AI Agent systems that abstract analyst workflows into production-ready automation

• KarrotMetrics Anomaly Reasoning Agent
  AI system for automated anomaly investigation across analytics alerts

  • Decomposed a 7-step analyst reasoning workflow into 6 MCP-based tools
  • Iteratively refined prompts up to v51 for more reliable reasoning
  • Reduced analysis time from 30 minutes → 1 minute (30× faster)
  • Deployed automated analysis across 13 alerts × 4 countries within 3.5 weeks

• KarrotMetrics LLM Report
  LLM-based reporting system for analytics teams

  • Designed separated contexts for daily/weekly/monthly reports
  • Improved hallucination and repetitive outputs using 8 prompting strategies
  • Adopted by 7+ teams within 5 weeks
  • Reduced manual analysis workload by 65–130 hours/month

• Experiment Platform Cache
  Caching system for analytics APIs

  • Improved API response time from 28.6s → 3.9s (7.4× faster)
  • Achieved up to 15.8× performance improvement

🍴 Woowahan Tech Course 7th · 2025.02 ~ 2025.11

Intensive software engineering program focused on OOP, TDD, and production systems

• ⚾ Yagubogu · Backend Engineer 2025.07 ~ Present
  Production baseball attendance verification platform (Android 3 · Backend 4)

  • Reduced crawling load by 88% using a 3-layer ETL pipeline with adaptive polling
  • Improved ranking query latency from 44.92s → 0.3ms through composite indexes and query refactoring
  • Reduced CI/CD build time from 408s → 27s
  • Solved concurrency issues using Rate Limiting and Idempotency Keys

• Tech Talk Speaker : Practicing OOP with Spring Framework · 2025.08

• Reliable Agent Systems
• Context Engineering for LLM Applications
• AI-assisted Developer Workflow
• Observability & Guardrails for AI-generated Code
• Production-grade AI Infrastructure