In this AWS Summit ANZ 2022 session, Sara van de Moosdijk (Moose) presents a comprehensive guide on designing end-to-end MLOps architectures for organizations of varying sizes and maturity levels. She walks through the challenges of deploying and maintaining machine learning models in production, then introduces MLOps concepts and AWS services that can address these challenges. Moose provides detailed architecture diagrams for small, medium, and large MLOps setups, explaining key components like automated retraining pipelines, model registries, and monitoring tools. The session aims to help architects and developers implement effective MLOps practices using AWS services like SageMaker, Lambda, EventBridge, and CodePipeline. Moose emphasizes that MLOps is a journey, encouraging viewers to start with basic steps and gradually adopt more advanced features as needed.