Skip to main content
Data extraction (OCR)
A
Written by Anna Dziurosz
Updated over a month ago

What is OCR?

Optical Character Recognition technology (OCR) is the process of extracting the written text in a document and finding its meaning. It can convert text (printed or handwritten) from documents into information in your system. This eliminates the need for manual data entry and streamlines your processes.

The OCR functionality is available in our webapp as well as both Android and iOS versions of our mobile app.


Accuracy

Our data extraction accuracy reaches up to 99% for most customers, requiring only 1 in 100 invoices to have extracted data manually corrected.

Accuracy can vary per invoice or receipt due to various reasons:

OCR can fail, when the provided images (in our case invoices) have a low resolution. If the provided invoice is handwritten, the OCR technology can have issues scanning certain types of handwriting. Other factors such as a complex layout and unusual font sizes can also affect the scanning abilities.

When is OCR technology being used

OCR technology is in use for various of our products, such as receipts, accounts payable and out of pocket.

Invoices

Invoices that are being sent to Moss are automatically being scanned by our OCR technology. We rely on OCR to automatically populate the invoices fields. Once OCR has successfully scanned the supplier in an invoice, we can automatically populate the accounting fields, saving time and avoid manual mistakes.

Receipts

We use OCR technology to extract the transaction number from your receipts automatically. This unique identifier is then stored alongside your transaction details in Moss, ensuring that your proof of payment is always accessible and easy to find. Whether you need to track expenses or validate a purchase, our OCR system simplifies the process, saving you time and keeping your records organized.

Reimbursements

Using OCR when submitting a reimbursement is remarkably straightforward. After accessing the reimbursement section within our mobile app, proceed to initiate a new reimbursement request. When prompted to attach supporting documents, select the option to upload a file. Choose the image containing the receipt or invoice from your device's gallery. Our OCR technology will then process the image, extracting pertinent details such as the vendor name, date of purchase, and total amount spent, and populate the corresponding fields in your reimbursement form automatically.

How do you know that OCR was used

You are able to identify that OCR technology was used by Moss when you see the purple stars next to your data fields. The purple stars mean that we scanned the document that you provided and automatically populated the fields.

What happens if the OCR process takes longer than expected?

In the event that the OCR process exceeds a predefined timeframe (typically 10 seconds), you will be presented with the option to cancel and proceed with manual data entry. This ensures that you maintain control over the reimbursement submission process and can opt for an alternative approach if necessary.

Did this answer your question?