Building services that customers actually need
Dohyeon Kim
Software Engineer → Technical Program Manager (8yrs+)
8yr Backend Expert Built and operated high-availability services in Java/Kotlin MSA environments across fintech and e-commerce.
Product-Oriented Drive decisions with data and UX insights, prioritizing business outcomes beyond code.
Cross-Functional Leader Led complex, multi-team programs end-to-end — aligning DBA, data, marketing, compliance, and engineering stakeholders.
Data-Driven Decision Maker Translate structured & unstructured data into actionable business insights and engineering priorities.
Global Communicator 6 years of overseas experience (Sweden). Proficient in English for business communication and cross-cultural collaboration.
Professional Experience
- Designed and built a Kafka-based notification delivery pipeline serving millions of customers across push, SMS, and KakaoTalk channels
- Led monitoring system decoupling — coordinated DBA, data, marketing, and dev teams to decouple Hadoop dependency, achieving 24/7 uptime and 80% response time improvement
- Resolved Kafka partition imbalance and introduced reactive consumers, improving TPS from ~400 to ~10,000 (25× throughput gain)
- Rebuilt 30M-record log pipeline with Project Reactor parallelization — processing time reduced from 10 hours to minutes
- Resolved Redis CPU 100% incident by replacing List (O(N)) with Set (O(1)) data structure
- Reduced pipeline creation effort by 80% — proposed and implemented a copy feature based on 1-year log analysis, aligning compliance and operations teams
- Led "My Document Box" integration without PM — achieving 100× subscriber growth (100→10,000/day) through scope negotiation and tech stack decisions
- Developed account-password-based certificate issuance — now ~60% of all certificate issuances
- Improved review service p90 API response from 250ms → 30ms and resolved DB connection pool exhaustion
- Designed and built Style Tab service backend (Clean Architecture + Multi-module) — achieved 30–40% revisit rate (2× industry average)
- Led exhibition system development as PM — designed strategy pattern architecture while managing stakeholders
- Proposed data-driven UX improvements based on Amplitude analysis; adopted by product team
- Designed and developed statistical price anomaly detection system (patented)
- Led retail (Rocket) product catalog service decomposition and system migration
- Designed product modification history architecture and built catalog management back-office
- Led legacy API replacement program end-to-end
- Alternative Military Service as Technical Research Personnel (Oct 2018 — Oct 2021)
Selected Projects 2024 — 2026
Program Management & Cross-Functional Leadership
System Re-engineering & Technical Execution
Tech Stack
Education
- Industry-academia projects with Hyundai Motors, Samsung Electronics, Samsung Heavy Industries
- Research focus: Data Mining, Process Mining, Anomaly Detection
- Thesis: Multi-perspective Anomaly Detection Using Process Mining