High-Performance Entity Matching Solution Using Vector Embeddings and AWS

Learn how to build a high-performance hotel matching system in this technical talk from the Vector Space Talks series. Explore a sophisticated solution that combines AWS infrastructure, Qdrant vector database, and vector embeddings to solve the complex challenge of matching hotel information across multiple sources. Discover the implementation details of key components including MiniLM-L6-v2 for generating embeddings, serverless inference with AWS Sagemaker, vector storage optimization using Qdrant, and Lambda with API Gateway for handling matching requests. Gain practical insights from real-world experimentation and understand how this architecture achieves improved speed and accuracy in hotel mapping, leading to enhanced customer satisfaction and operational efficiency at HRS Group. Follow along as Data Engineer Rishabh Bhardwaj shares his expertise in developing scalable and cost-effective solutions for the travel industry using modern vector search technologies.