Q: We receive an increasing number of invoices via email as attachments, do we have to print these and then scan – it sounds like a lot of work?

No, FIS/fci can take the attachments and process them without the need for scanning. This is an ideal solution because invoice data can be extracted without any human intervention.  In addition, the invoices are normally created directly from an ERP system and this makes the reading of the invoice data easier as fidelity and accuracy is not lost through the scanning process.

Q: What about recognition rates, how many invoices need manual corrections?

There can be a large variance in recognition rates and this is largely down to factors such as whether header (i.e. supplier, PO number, invoice number, total net etc) or lines (description, unit cost, number of units, net cost of line etc) are required. Typically rates of 90% + are common. The knowledge database always improves the recognition rates from the initial free-form recognition methods. FIS/fci uses a form of machine learning – becoming more accurate at reading data as the number of invoices increases.

More info

Q: The data on the supplier invoices is different for every supplier?

FIS/fci is designed to search for the required invoice data for every supplier – irrespective of the layout. The “Freeform Recognition” software works for of different invoice layouts for each supplier.

Q: How does the software know that the document is an invoice?

One of the first checks the software performs is to identify if the word invoice (or a translation of the word invoice – facture, fattura, faktura, rechnung etc) is present. The system is also able to recognise different company codes by a look-up of a reference file from the ERP system.

More info