Dongyoung Lim Email & Phone Number

South Korea

South Korea

[Channel Server Part]📌 Kapayback: Scalable Payback Service Development •Developed a scalable payback system leveraging user payment history. •Integrated Kafka to handle high-traffic loads asynchronously, reducing API requests by 90% and significantly enhancing system reliability under heavy traffic conditions. •Proactively collaborated across multiple teams (payment platform, OLAP platform, IFTTT-like platform, reward platform) to design and implement a robust and seamless payback system.📌 Offline-First Payment Promotion Server •Designed a system to reward users who had not used the offline payment service in 2024. •Maximized utilization of existing internal platform APIs, minimizing resource consumption for this promotion and eliminating the need for platform department development efforts.📌 Push Message Middleware Kafka Consumer •Developed a Kafka consumer to throttle TPS for integrated APIs and Message APIs, preventing peak traffic overload when sending push notifications. •Implemented an extensible structure using the strategy pattern to efficiently handle various types of messages.

[Hub Server Unit]📌 Developed Kakaopay Home API server focused on performance and scalability •Utilized Spring Boot, Kotlin Coroutines, and R2DBC to implement non-blocking IO and concurrency, reducing p90 latency from 2 seconds to 500ms (a 75% improvement). •Designed Server-Driven UI, cutting API versioning complexity and reducing native client release dependencies, which removed dependency on update cycles occurred by app releases.📌 Developed a user payment-based donation promotion service for Christmas 2023 •Designed a system to dynamically respond to user payment events using an internal IFTTT-like platform, triggering donation promotions in real-time. •Utilized Redis to aggregate user donation data, reducing database queries by 95% (from 20 queries per minute to just 1 query per minute).

대한민국 서울

📌 Migrated Moreden's backend from Python Flask to NestJS and AWS ECS Fargate •Successfully handled traffic surges of over 10,000 users without service downtime during high-traffic promotions with ECS Fargate containerized service (migrated from a single t2.medium EC2 instance to two 0.5 vCPU / 1GB RAM Fargate instances, scaling out to 8 instances during peak periods).📌 Revamped Cheestalk's full-stack system to improve development efficiency •Reduced meeting time by 66% through streamlined communication through agile scrum sprints. •Utilized BDD (Behavior-Driven Development) to improve alignment between product requirements and implementation, facilitating smoother communication with product managers and ensuring the tests accurately reflected business logic.📌 Developed a job matching system apply form page for dental hygienists using React.js, AWS Amplify, and Lambda within 5 days, speeding up deployment cycles.

📌 Developed an internal tool for data inspection, improving workflow efficiency: •Utilized Python and Qt to create a GUI for XML data management in Perforce. •Automated data modify processes, reducing errors and improving team productivity.