Executing machine learning (ML) models at the speed and scale of production is a key challenge as enterprises continue to operationalize artificial intelligence. SingleStore, formerly known as MemSQL, provides a converged operational database that combines capabilities found in multiple datastores, such as documents, key-value pairs, stream processors, and relational databases, into a single data platform to activate your most critical data for operational BI and data science. In this talk, Engineering Manager Carl Sverre discusses how SingleStore leverages AWS services in the ML workflow to compile and run Amazon SageMaker models as database functions against real-time data.