Projects
Kubernetes and RabbitMQ Integration (CareSquare)
• Stacks: EKS, Helm, Karpenter, RabbitMQ, EFS, Docker
• Migration & Scaling: Led the migration of microservices, APIs, and web servers from EC2 to EKS. Utilized Helm
and Karpenter for dynamic autoscaling, which optimized resource usage and reduced server costs by 50%.
• CI/CD Automation: Developed a CI/CD pipeline to automate builds, tests, and deployments, enhancing de-
ployment reliability and reducing deployment time by 80%. This automation minimized manual tasks and
errors, significantly boosting overall productivity.
• Logging & Monitoring: Created a centralized logging system with EFS and RabbitMQ, improving diagnostics
and reducing incident response times (MTTR). Enhanced operational oversight and troubleshooting eiciency.
• Performance Optimization: Implemented autoscaling and health checks across Kubernetes services for high
availability. Optimized Docker containers with Alpine images, enhancing performance and reducing resource
use.
Chronic Disease Care Project (CareSquare)
• Stacks: EKS, RabbitMQ, Go, Echo
• Cost Reduction: Developed a custom AlimTalk client, significantly reducing messaging costs by 50% com-
pared to third-party services, optimizing communication eiciency across the platform.
• Automation: Implemented Cron jobs and developed workers to automate the delivery of AlimTalk notifica-
tions and surveys, eiciently scaling the system to serve tens of thousands of patients weekly.
• API Integration: Engineered custom clients for seamless integration with Lotte Information & Communication
APIs, utilizing Goroutines to enhance API call eiciency and reliability across multiple hospital systems.
• Survey System: Built the survey API server, automating key processes in patient management and response
handling, thereby improving overall operational eiciency on EKS.
Image Crop Tool (CareSquare)
• Stacks: AppScript, AWS Lambda, PHP, Nginx, OpenCV, Python
• Achievements: Automated the cropping of 30,000 images in one month, cutting processing time by 70%
and enhancing the delivery of educational content.
• Automated Cropping: Used OpenCV to develop a system that detects and crops image boundaries, eliminat-
ing manual cropping and increasing throughput.
• Web Interface and Distribution: Created a Nginx-based web interface for verifying crops and employed AWS
Lambda to automatically compile and distribute images to Google Drive.
Landing Page (CareSquare)
• Stacks: EKS, ImageProxy, RabbitMQ, Go
• Achievements: Increased user acquisition by 30% through improved targeting and funneling from multiple
platforms. Reduced image load times to 10% of the original, enhancing user experience and cutting network
costs.
• Optimization and Tracking: Implemented user flow tracking with VWO and deployed an image cache server
using Image Proxy on Kubernetes, optimizing image handling. Integrated RabbitMQ for eicient data flow and
logging.