Documents & OCR

VegaERP stores documents centrally and can extract structured data from them automatically — so a scanned supplier invoice becomes a draft expense instead of a manual data-entry task.

Uploading documents

Upload a document directly, forward it by email if you’ve connected your inbox, or receive it automatically through a connected Gmail/Outlook account — anything landing in the unified inbox that looks like an invoice or receipt is picked up the same way.

What OCR extracts

For supported document types (mainly supplier invoices and receipts), OCR processing extracts:

• Supplier identity (name, tax number) — matched against your existing contacts where possible.
• Invoice number, dates, and amounts (net, VAT, gross) per line and in total.
• VAT rate per line, where legible.

The result is a draft document for you to review and confirm, not an auto-booked entry — OCR extraction can misread a scan, and a wrong amount or VAT rate booked automatically would be a compliance problem, not just an inconvenience.

Reviewing extracted data

Always check, before confirming:

• The matched supplier is actually the right contact (a similar company name can match the wrong existing contact).
• Amounts and VAT rate match the source document, especially on lower-quality scans or photographs.
• The document date and posting period are correct.

Multi-warehouse and inventory documents

Delivery notes and other inventory documents follow the same upload → extract → review flow, feeding stock movements once confirmed rather than requiring separate manual stock entry.

If extraction quality is poor

Low-resolution scans, skewed photographs, and handwritten amendments reduce OCR accuracy. Re-scanning at higher resolution or straightening a photo before upload usually fixes repeated extraction errors on the same document type.