A company wants to run ML analytics on data in an S3 data lake. They have two transformation needs: (1) perform scheduled daily transformations on 300 GB of various-format data that arrives at a set time, and (2) run a one-time transformation on terabytes of archived data in the lake. Amazon MWAA DAGs will orchestrate the workflows. Which two tasks should the DAGs schedule to meet these needs most cost-effectively? (Choose two.)

AWS Glue crawlers are the most cost-effective solution for automatically scanning and inferring schemas from various data formats in S3, making them ideal for the scheduled daily transformations. Amazon EMR is a powerful and flexible big data processing service that can handle both the scheduled daily transformations on 300 GB and the one-time transformation on terabytes of archived data. Its ability to scale and process large datasets efficiently makes it cost-effective for these diverse needs. Amazon Athena is a query service, not designed for schema inference, making it unsuitable for scanning and identifying schemas. Amazon Redshift is a data warehouse, optimized for analytical queries on structured data, not for general-purpose data transformations on varied formats in a data lake. Amazon SageMaker is primarily for machine learning model training and deployment, not for large-scale data transformation tasks.