A company must ingest large volumes of data from many sources, apply multiple business-rule transformations in sequence, support reprocessing when errors occur, be scalable, and require minimal maintenance. Which AWS service should be used to orchestrate and automate these data flows?

AWS Step Functions is the correct choice because it allows you to orchestrate complex workflows as a series of steps, making it ideal for sequential data transformations and error handling (reprocessing). It's inherently scalable and requires minimal operational overhead. AWS Batch is for running batch computing workloads, not primarily for orchestrating sequential data flows. AWS Glue is an ETL service, excellent for data transformation, but Step Functions provides the overarching orchestration and state management needed for complex, multi-step workflows with reprocessing. AWS Lambda is a serverless compute service for individual functions, not designed for orchestrating multi-step, stateful workflows.