During invoice processing, invoices that arrive in PDF format via the corresponding interface are processed automatically. In this process, the company is first recognized, then a distinction is made between invoice and credit memo and finally the invoice is indexed.
Automatic company recognition
A company is preset by default based on system information from the document service.
However, it may also be the case that several companies come into question for an invoice (e.g. the customer has several addresses, branches, etc.). In this case, the appropriate company is automatically selected based on system information and the information available on the invoice.
Distinction between invoice and credit memo
Each invoice is checked using various criteria to determine whether it is an invoice or a credit memo. All subsequent processes (e.g. recognition of the invoice number and credit memo number) are set according to the identification.
Automatic keywording and vendor recognition
During keywording, the invoice is checked for various criteria. The vendor is automatically recognized in this process.
The process searches specifically for the following entities:
-
Purchase order numbers
-
Gross amount
-
Creditor (address, IBAN, name, tax ID, VAT ID)
-
Delivery numbers
-
Net amount
-
Line items
-
Invoice date
-
Invoice number
-
Tax amount
-
Tax rate
-
Tax ID
-
VAT ID
-
Currency
If the vendor has not been recognized, the process searches for a city name in order to automatically perform a comparison with a database of city names. Possible creditors are then loaded based on the zip code. The correct creditor is then used based on various mechanisms (matching of names, IBAN, etc.).
The database of city names only includes cities in Germany and Switzerland (as of 2021). Please get in touch with your edoc contact if you require an extension to the database.
Adopt already recognized entities
Of course, you can also have already established your own recognition process for incoming invoices. A human or a reader already recognizes and defines the entities of the invoice. In this case, edoc document reader ignores the entities that have already been recognized. This ensures, for example, that existing entities are not overwritten.