This is a dedicated watch page for a single video.
You recently migrated the backend data system for an analytics application used by NovaRetail Inc., which manages 10 PB of historical product data. The system must support analytics on all historical records, but only the latest product state—about 10 GB in total—needs to be served via an API to other internal services. The API must handle up to 1000 queries per second (QPS) with sub-second latency. You need to select a persistent and cost-efficient storage architecture that supports both large-scale analytics and high-performance API responses. What should you do?