In this talk, I'll demonstrate how to create a Retrieval-Augmented Generation (RAG) application on AWS. The application will be backed by AWS services such as Bedrock, Lambda, S3, the vector database LanceDB, and FastAPI. During the talk, I'll also present some tips on different approaches to save costs when creating a RAG system with low traffic and how to optimize performance. You'll learn practical techniques to ensure your application is both cost-effective and scalable, making the most out of AWS's serverless infrastructure.