User-Perceived Backup/Restore Success Rate Metrics

A backup/restore operation only succeeds when a chain of server calls all complete — I defined and automated a metric for that end-to-end, user-level success

• Moved from per-API success checks to a success rate defined over user-operation transactions spanning the full call flow
• Automated metric extraction with Airflow/Python and daily-refreshing Grafana dashboards, plus auxiliary metrics for failure-cause tracking
• Drove improvement actions — success rate ~70% → 95%

Success rate ~70% → 95%

SLI/SLO Anomaly Analysis & Reporting Automation

Replaced manual log digging and chat-based reporting with an AI agent workflow

• Automated extraction of analysis conditions (service, region, API, time window) from alert payloads, log collection, and request/trId-level clustering
• Chained a root-cause-analysis agent and a report agent posting standardized summaries
• Analysis-to-report time: up to 4 hours → ~10 minutes

Analysis-to-report time: 4h → ~10min

Service Runtime Modernization (Java 25 · K8s · Graviton)

Laid the groundwork to migrate build, CI, and metrics architecture in one coordinated move

• Updated build/test tooling for Java 25; refreshed CI workflows and Docker base images
• Migrated K8s metrics to OTLP (OpenTelemetry); set up ARM configs for Graviton validation
• Analyzed AWS SDK v1→v2 timeout and DynamoDB behavior differences; corrected acceptance tests

"AgentHub" — AI Agent Session Manager

One screen for the status and control of scattered AI agent sessions — solo, from planning to deployment

• Built tmux session collection, status badges (running / waiting / approval needed / error), web terminal attach
• Implemented as an Electron/React desktop app and shipped to the internal Tool Store
• Ranked

Ranked

Also

• Expanded E2E tests for core backup/restore scenarios; removed flaky tests with embedded Kafka
• Built a team Jupyter/PySpark log-analysis tool; automated weekly work-logs via GitHub + LLM summarization

DOWITH

Collaborative habit-building service (Pirogramming)

MOA

Dorm delivery-mate matching system (GDSC)

CPS

Classroom intrusion detection system, tied to my industrial security major

SSAFY Attendance Excuse Generator

A utility with real users

Booster

Print-ahead app concept

Chung-Ang University

B.S. in Industrial Security & B.S. in Software (double major)

Graduated with honors · GPA 4.14/4.5 (95.89/100)

• CS: Data Structures, Algorithms, Operating Systems, Computer Architecture, Software Engineering, Programming Languages, Network Application Design, Machine Learning, Probability & Statistics, Software Project
• Security: Information Security Theory, Security Data Analysis, Cyber Incident Response, Industrial Security Investigation & Forensics

Korean (native) · English (advanced) · Chinese (intermediate, business communication)