PRODUCTION SCENARIO
An insurer stores 250,000 scanned claim forms as PDFs. An analyst wants average claimed amounts by region, but the amounts and addresses exist only inside the scanned pages.
How should the data be classified, and what must happen before the report can be built?
Answering here is anonymous. Nothing is saved unless you sign in.
Show answer and explanation
Answer: Unstructured data; the fields must be extracted into structured records first
Document processing takes unstructured content and turns it into structured data, meaning named fields suitable for a database. Until the amounts and addresses exist as fields, there is nothing for a query to aggregate by region.
Official Source: https://docs.cloud.google.com/document-ai/docs/overview ↗