Building a Robust Feature Platform for Data-Powered Recommendations

Explore a Korean-language lightning talk showcasing how Karrot, a Korean hyperlocal community service, built a sophisticated feature platform to enhance its recommendation system. Dive into the architecture that leverages Amazon EKS, Amazon MSK, Amazon DynamoDB, Amazon ElastiCache, and AWS Batch to streamline data processing workflows and enable large-scale data operations. Learn about the integration of generative AI for extracting actionable insights and discover how this robust platform improves the user experience on Karrot's mobile application through efficient data collection, processing, and serving capabilities.