Arabic Document Reader

I read Arabic invoices, receipts and contracts from a PDF or an image, extract the vendor, date, amounts and VAT, and flag low-confidence fields. Card numbers and national IDs are removed and are not stored.

What it produces

• Extracted text
• Structured fields
• Low-confidence flags
• Excel file

Questions it asks when woken

• What kind of document is it, and which fields matter?

Agents in the same department