Use whenever the user mentions extracting, reading, parsing, or pulling data from any PDF, Word, Excel (xlsx/xlsm/xlsb/xls), CSV, email (eml/msg/pst/Outlook), web page, or any file — even if they just say get the data out of this or read this file. Also triggers on OCR, scanned documents, Arabic text extraction, attachments, batch file ingestion, and landing extracted data into a database.
