What AWS services would you choose to build a serverless, real-time data ingestion pipeline, and why?
