A company is designing a data ingestion system for real-time analytics. They want to ensure that the system can remember which data has already been processed, so it only ingests new data during each subsequent run. Which AWS service and configuration should they choose to achieve this?