PDF Document Parsing Detailed by Line Item

The eWorkflow OCR for ePosting enables several ways of reading PDF documents or invoices. The standard ePosting functionality currently supports two methods for reading PDF documents.

One of the two methods of reading a PDF document is detailed by line item (line-item-invoice).

To help you understand this, a use case was prepared, in which an accountant is processing a document received from the supplier. This document has four lines. The reading the PDF document detailed by line item results in an invoice received in PANTHEON, which has four lines, just like the supplier's source document.

The example below describes the above process in more detail.

CASE SUMMARY The person responsible wants to import an invoice for the subject Maribor Adult Education Centre using the PDF parsing method detailed by line item. This is done as follows: Defining/checking the PDF document parsing method Running the eWorkflow OCR for ePosting Importing the received e-invoice Completing the missing data for ePosting machine learning Checking the received invoice in PANTHEON

1. Defining/checking the PDF document parsing method

First, the person responsible defines or checks the PDF parsing method setting for the selected subject in the Subjects register.

In this case, the line-item-invoice option must be selected in the Data column, as shown in the image below.

HINT For more detailed instructions on how to set up the parsing (reading) method, see the chapter Setting up the method of PDF parsing.

Next, the person responsible imports the received invoice.

2. Running the eWorkflow OCR for ePosting

The person responsible imports the document into eDocumentation, as shown in the image below.

To convert the data from the PDF document into the e-invoice format, in the toolbar of the Documents window, the person responsible clicks the (Wizards) button and selects the OCR for ePosting option from the drop-down menu.

HINT To display the eWorkflow as an option in the Wizards menu on the specified PANTHEON form, you need to enable this in Workflow Design. For more information, see section 3 of the chapter Defining basic settings for eWorkflows.

The eWorkflow converts the PDF document into an e-invoice in JSON format, which is displayed in the Files tab, while the preview of the JSON file is shown in Document preview.

Next, the person responsible imports the e-invoice into PANTHEON.

3. Importing the received e-invoice

In the toolbar of the Documents window, the person responsible clicks the (Wizards) button and selects the eWorkflow to be run: Import received e-invoice to ePosting.

HINT For a detailed procedure describing how to import the received e-invoice into PANTHEON, see the chapter Importing the received e-invoice into PANTHEON using ePosting (OCR).

On the Import eDocument Receiving Documents form, the person responsible checks and completes the missing data.

HINT Some data might be read directly from the invoice without any difficulties, while the reading of other data will sometimes need to be additionally processed by PANTHEON logic.

4. Completing the missing data for ePosting machine learning

When importing an e-invoice, some data, such as a date or an invoice line, may not be read correctly. The system learns with every import and will be more successful the next time.

In this case, the system has not recognized some dates, external IDs, etc. Therefore, the person responsible enters the required corrections, as shown in the image above and explained below:

• In the header: OCR reader did not recognize the VAT date of the invoice, which is why the person responsible manually enters this data.
• An external ID is recorded on the e-invoice. In the first two lines of the invoice, PANTHEON identified its own (internal) ID based of the external ID and the subject. The settings for linking the external ID and the PANTHEON ID can be accessed in the Items register.

In this case, for the second and third line, the person responsible manually selects the ID from a drop-down list, as shown in the image below.

PANTHEON IDs can be recognized by the program in different ways. By checking the checkbox Add missing external IDs to Items register at import, the external IDs will be automatically added to the Subjects register upon import.

HINT When you want to import an item with the same external ID, PANTHEON will automatically recognize its internal ID.

If no unit of measurement is recorded on the e-invoice, the default unit of measurement from Exchange settings is automatically filled in.

The person responsible can now import the invoice. To select the document, the person responsible checks the checkbox in the beginning of the line, as shown in the image below, and click the Import button.

The invoice is created based on the Document type selected in the checkbox in the section in the As Invoice Date, VAT is applied section below.

The number of the imported document is displayed on the Links tab of the Documents window.

Double-clicking the document number opens the invoice window.

5. Checking the received invoice in PANTHEON

The received invoice, which was created by importing this e-invoice to PANTHEON, contains as many lines as there were lines in the PDF invoice.

HINT After an e-invoice has been imported into PANTHEON, if you notice that the OCR system has not read all the data correctly, e.g. dates and/or tax rates, you can manually update or change such data. The OCR system learns from the feedback submitted when the invoices are imported into PANTHEON. This means that it will be more successful the next time it reads them (ePosting machine learning).