Our solution will interpret the OCR results of an invoice in a way that all invoice features of invoices, such as its number, net and gross amounts, tax amounts for different VAT rates, date of payment, number of account of the supplier – i.e. all the data necessary for its decreeing – will be found. These data will be automatically entered in an assignment form. For good-quality scans, this method provides a very high probability of reading correctness.

Furthermore, we offer you an application for local use, not a cloud service, which seems to be particularly important when it comes to the security of sensitive data concerning the invoice value, subject, and the contractor.

The application is a solution that does not require reading templates modelling; it creates a library of invoice patterns and improves the details of templates in the process of subsequent data readings. Reading the data from the invoice unknown to the system requires an indication of those areas on the invoice scan from which the read numbers, phrases, and dates should go to the appropriate form fields. This operation for the “new” invoice is implemented only once, while allowing the application to learn a new invoice scheme. The use of our product allows to reduce the work of employees entering the invoices into the system to a minimum.

After applying the Invoice Description Automation product, working with repeating invoices looks as follows: after scanning, the role of the person entering the document into the system is limited to validating the correctness of phrases and data transferred to the form.