Invoice OCR Intake

OCR (Optical Character Recognition) lets you photograph or upload a supplier invoice and have FlexotiumPOS extract all the key data automatically. This eliminates manual data entry for the majority of invoices.

How to capture an invoice

Method 1 — Upload a PDF

  1. Go to Purchasing → OCR Intake → Upload
  2. Drag and drop or select your PDF file (max 20MB, up to 10 pages)
  3. Click Process — extraction takes 5–15 seconds

Method 2 — Photograph with mobile

  1. Open FlexotiumPOS on your smartphone
  2. Go to Purchasing → OCR Intake
  3. Tap Take Photo — the camera opens in document mode
  4. Frame the invoice so all four corners are visible
  5. Tap the shutter — the image is uploaded and processed automatically

Method 3 — Import an image file

Upload a JPG, PNG, or TIFF of a scanned invoice via Upload → Image File. Scans at 300 DPI or higher produce the best extraction accuracy.

What OCR extracts automatically

| Field | Extraction accuracy | |-------|-------------------| | Supplier name | ~95% | | Invoice number | ~98% | | Invoice date | ~96% | | Due date | ~90% | | Line items (description, qty, unit price) | ~88% | | Total amount | ~97% | | Tax / VAT amount | ~92% |

Reviewing and correcting OCR output

After processing, the extracted invoice is shown side-by-side with the original document for comparison. Fields that the system is uncertain about are highlighted in amber.

Correct any errors by clicking the field and typing the correct value. The system learns from corrections over time to improve accuracy for your specific suppliers.

OCR extraction matches line items to your product catalog using product names, descriptions, and supplier codes. Unmatched items appear in red and must be linked to a product or created as new before the invoice can be submitted.

Submitting after review

Once all fields are confirmed, click Submit for Approval. The invoice enters the normal approval workflow from this point.