2020.05 – Present · Place Data Platform (org leader since 2021.04)

Leading a 6–8 person data engineering org — transformed manual data request handling into a contract-based self-service platform, and now building a dev culture where AI agents work alongside the team.

AI Engineering & Dev Workflow Automation

• Led LLM-agent (Claude Code) based team automation — built and operate an AI code review bot, automated Jira issue triage/operations, and an internal AI assistant (Slack bot) on Kubernetes.
• Designed AI contribution measurement infrastructure — deterministic tracking signatures aggregated via LLM judgment: 400+ merged AI-authored PRs and 1,200+ AI-driven tasks measured within six months of adoption (as of Jul 2026).
• Standardized repetitive work (data modeling → pipeline contract → access requests) into reusable AI skills/plugins shared by the whole team.

Data Lakehouse

• Designed and operate an Iceberg·Spark·Kafka·Airflow data lakehouse — a ~90-node dedicated Hadoop cluster with a 100TB+ snapshot layer, realtime CDC from 13+ service sources, and active/standby disaster recovery.
• Led a zero-downtime IDC migration (2026) — moved 100TB+ of data and every component with a dual-write / fade-in-out strategy, cutting hardware by ~25%.

Place BI — Dimensional Modeling & Semantic Layer

• Established dbt-based dimensional modeling conventions (fact/dimension/serving layers) and built a static validation tool that auto-detects violations — model quality guaranteed by the system, not by people.
• Standardized ~100 metrics into a single source of truth with a MetricFlow semantic layer, delivered as KPI dashboards — eliminating cross-team data inconsistencies.
• Built an automated parity-verification harness that cross-checks KPI numbers produced by different systems, securing metric trustworthiness.

Data-Contract-Based Self-Service ETL

• Designed a self-service data platform where schema definition (Protobuf), pipeline creation, and data subscription/permissions are all handled by a single contract-file PR — operating 500+ data models, 30+ pipeline contracts, and 160+ dbt models.
• Supplies data to company-wide AI workloads — foundation-model pretraining, abuse detection, credit-scoring models — through 12 data products (DaaS).

Data Governance & Data Mesh

• Led data governance for the Place data domain and headed a Data Mesh transition TF, designing a domain-oriented data ownership structure.

Realtime Abuse Detection

• Advanced a realtime-pipeline-based abuse detection system, raising service trustworthiness. Tech blog (KR)

2018.05 – 2020.05

Turned an unproven technology — blockchain — into production-grade infrastructure.

• Researched and developed blockchain transaction performance improvements and interchain systems.
• Built KMS/HSM-based secure blockchain network provisioning and Docker/Kubernetes node orchestration. Tech blog

2016.10 – 2018.05

Built the system absorbing a global messenger's ad traffic — recognized as best engineer for it.

• Developed a high-performance ad event collection system handling 2B daily requests / 4M TPS (Kafka, Avro, ELK, Hive).
• Developed ad-serving servers and an admin permission-management module (Spring, Scala, Akka-HTTP).
• Received the Best Engineer Award (2017) for outstanding technical contribution and business impact.

2011.01 – 2016.10

Community, search, music streaming — built the fundamentals of large-scale systems across global services.

• Designed and developed the LINE MUSIC streaming service, plus its settlement and billing (IAP/IAB) systems (Spring, MySQL, ELK, Spring Batch).
• Developed the LINE messenger Note service — high-volume processing on sharded MySQL.
• Built LINE Thailand shopping search (Elasticsearch) and Naver Japan Cafe services.

2008.07 – 2010.12 (internship 2008.03 – 06)

From intern to Rookie Award in year one, Excellent Employee Award the next — proving growth speed from the start.

• Led game logic and backend server development for 'Hoyfarm', the predecessor of the hit social game 'Rule the Sky'.
• Built an in-house Lucene-based search engine and web crawler, improving data collection and search efficiency.
• Converted to full-time after the 2008 internship; won the Rookie Award and Excellent Employee Award in consecutive years.