How would you design a scalable data pipeline for ingesting and processing clinical data at nference?
