A retailer stores 100 GB of transactional data in S3 daily and needs to detect the data schema, perform transformations on the S3 data, and use ML to detect fraud, all with minimal operational overhead. Which combination of services should they use? (Choose three.)

AWS Glue crawlers are ideal for discovering schema from data stored in S3, automatically cataloging metadata in the AWS Glue Data Catalog. This addresses the need to detect the data schema with minimal operational overhead. AWS Glue workflows and AWS Glue jobs provide a serverless, scalable way to perform ETL (Extract, Transform, Load) operations on data, which is perfect for transforming the S3 data without managing infrastructure. Amazon Fraud Detector is a fully managed service specifically designed for fraud detection, offering pre-built models and expertise to train a fraud-detection model efficiently. Amazon Athena can scan data and infer schema, but AWS Glue crawlers are more robust for persistent schema discovery and cataloging. Amazon Redshift stored procedures are for data within a Redshift cluster, not directly for transforming data in S3. Amazon Redshift ML can train models, but Amazon Fraud Detector is a specialized service for fraud, offering a more streamlined and effective solution for this specific use case.