Document + Email Data Extraction

Our system processes documents (PDFs, emails, scanned files) and uses AI to extract structured data into records. Each record contains fields like dates, amounts, names, codes, etc.

Your job is to **review every extracted record**, comparing the AI’s output against the source document, and:

- **Confirm** fields that are correct
- **Correct** fields that are wrong
- **Fill in** fields the AI missed
- **Flag** anything ambiguous for escalation
⚠️

**Important:** A single job may include **various document types** (invoices, contracts, bills of lading, certificates, etc.). The extraction schema (i.e., the fields you review) will change depending on the document type. Always check the field names and descriptions carefully — do not assume fields mean the same thing across different document types.