The article discusses the implementation of a large-scale vector database designed for real-time recommendation systems in the LINE VOOM platform. It highlights the technical challenges faced and the solutions developed to enhance user experience through personalized content delivery. The focus is on leveraging advanced database technologies to improve the efficiency and accuracy of recommendations.