Google Certified Professional Data Engineer: Deploying Machine Learning Pipelines — Flashcards

You have been tasked with helping to establish ML pipelines for your department. The models will be trained using data from several sources, including several enterprise transaction processing systems and third-party data provider datasets. Data will arrive in batches. Although you know the structure of the data now, you expect that it will change, and you will not know about the changes until the data arrives. You want to ensure that your ingestion process can store the data in whatever structure it arrives in. After the data is ingested, you will transform it as needed. What storage system would you use for batch ingestion?