The document discusses building recommendation engines using Amazon SageMaker. It begins with an overview of why companies build recommendation engines and common techniques like matrix factorization. It then introduces Amazon SageMaker as a service that allows users to easily build, train, and deploy machine learning models. The document demonstrates how to develop, train, and deploy a recommendation engine in Amazon SageMaker in 15 minutes. It concludes with examples of how some customers have used Amazon SageMaker to build recommendation systems for applications like ecommerce and content suggestions.