Which solution will meet these requirements with the LEAST operational overhead?
Use AWS Glue to create an ML transform to build and train models. Use Amazon OpenSearch Service to visualize the data.
Use Amazon SageMaker to build and train models. Use Amazon QuickSight to visualize the data.
Use a pre-built ML Amazon Machine Image (AMI) from the AWS Marketplace to build and train models. Use Amazon OpenSearch Service to visualize the data.
Use Amazon QuickSight to build and train models by using calculated fields. Use Amazon QuickSight to visualize the data.
Explanations:
AWS Glue is primarily for ETL (Extract, Transform, Load) tasks and is not specifically designed for building and training ML models. Amazon OpenSearch Service is useful for search and analytics, but it does not offer the full capabilities required for integrating ML models with reporting.
Amazon SageMaker is a fully managed service that enables building, training, and deploying machine learning models. It integrates seamlessly with Amazon QuickSight, allowing for easy visualization of the augmented data in business intelligence dashboards, thus minimizing operational overhead.
While using a pre-built ML AMI might simplify some setup, it lacks the integrated environment and automation that Amazon SageMaker provides. Furthermore, Amazon OpenSearch Service does not directly support ML model training or integration for reporting and visualization needs.
Amazon QuickSight is primarily a business intelligence tool for data visualization and reporting. It does not have the capabilities to build and train ML models effectively. Using calculated fields for ML model training is not a suitable approach and would not meet the requirements outlined.
I draft that the answer is:
Use Amazon SageMaker to build and train models. Use Amazon QuickSight to visualize the data.