OCR, InvoiceSharing and the bot user
PDF invoices via OCR
It is possible to work with the recognized OCR data in Onventis. The recognised OCR data can be identified by the blue framework. These frames are standard visible in the worklist when the transaction is still in the first step of the entire flow. To show the frames of recognised data manually or to hide them, click on the three dot colon and subsequently on Show OCR regions:
Besides that, the data can have various coloured frames for different meanings:
Blue: To indicate all recognised data.
Green: To indicate the recognised data based on the field you are positioned at on the transaction form.
Orange: To indicate that the manually entered data in the transaction form does not match with the OCR recognised data.
The recognised OCR data are always filled in, even when the supplier is not recognised.
It is important that the document is of good quality. When it is scanned skewed or has a coloured background, the chances on high quality result will be lower. Also handwritten notes on document can affect the quality of recognized data.
For recognizing data, the OCR looks for certain key words on the invoices such as ‘ invoice number’. The data next to this, will be considered to be the invoice number. The supplier is ‘ found’ by looking for the bank account number, Chamber of Commerce number or VAT number.
(E-)Invoice formats
Organizations receive invoices in various forms and formats, ranging from modern (hybrid) e-invoices via the Peppol network to traditional paper letters. Onventis Invoice Processing can handle all these scenarios. Let's explore the most common invoice formats:
PDF invoices
The most common way to receive an invoice is in PDF format via email. You can automatically extract these email attachments from your mailbox and process them in Onventis. Onventis uses enhanced OCR functionality to scan and recognize data on the PDF invoice, even at the line-item level. Customizations to the scanning process can be made, and if recognition fails for any reason, an OCR optimization request can be created.
(Hybrid) E-invoices
E-invoices come in various standards, such as XML UBL 2.2, XML ZUGFeRD, XRechnung, and others, depending on the region. Sometimes, a PDF file is embedded in the e-invoice. Onventis can handle many of these e-invoice formats and process them in Onventis Invoice Processing. The biggest advantage of a structured e-invoice is that the data is always 100% correctly interpreted, allowing for faster and more cost-effective invoice processing.
Peppol
Peppol is a secure network for sharing e-invoices, ensuring that both the sender and receiver are verified. Peppol invoices are inherently e-invoices and come with all the benefits that e-invoices provide.
Conclusion
Thus, whether you're dealing with PDF invoices, various e-invoice standards, or secure Peppol transactions, Onventis Invoice Processing ensures efficient and accurate processing. By leveraging advanced OCR technology and supporting multiple e-invoice formats, Onventis helps organizations save time and money while improving the accuracy and speed of their invoice processing workflows.
If your current configuration does not support one or more of the invoice formats mentioned, please contact your account manager to explore the available options.
Processing invoices
Now that we understand the different invoice formats, let's explore how invoices are processed and how the Onventis Hybrid Artificial Intelligence setup can help organizations save time and money.
Coding dimensions based on a purchase order
Regardless of the invoice format, if a purchase order is linked, matching can often be done automatically using a three-way match. This is highly beneficial as the purchase order already indicates the relevant financial dimensions (cost center, general ledger, etc.) and confirms receipt.
Coding dimensions based on artificial intelligence
When a purchase order is unavailable, the process becomes more complex. In such instances, we leverage Artificial Intelligence to generate a coding proposal. This proposal is derived from various factors, including invoice details, historically booked invoices, and your financial dimensions. This approach ensures that invoices are consistently coded, even without a purchase order.
Coding dimensions based on supplier defaults
In some cases, specific suppliers deliver only certain goods that can always be booked under the same financial dimensions. For these scenarios, the AI-generated coding dimensions can be disregarded, and the default booking dimensions can be used instead.
Summary
In summary, Onventis Invoice Processing typically relies on purchase orders, AI-powered coding proposals, or supplier defaults to determine the coding of invoice lines. Whether the invoice is from a one-off supplier or a frequently used one, our various methods can always provide you with an accurate coding proposal. This will save you time and money while enabling faster invoice payments.
This article explains the invoice processing methods available. Many existing customers have specific setups in place. To enhance your current setup, such as adding AI coding proposals, a consultant will need to configure it. Your account manager can discuss the available options and license costs with you.
Automatic invoice processing
The bot user can take over entering and matching an invoice.
The bot user used the data of OCR or, even better, the digital invoices from InvoiceSharing. Based on available data, the bot looks for matches in receipts or orders, or uses transaction defaults to enter the lines.
The bot checks the invoice for differences. If a deviation is found, the bot user will forward the transaction to the worklist of the responsible user to deal with the transaction. If no mismatch can be found, the bot can process the invoice without further interaction.
The bot is separate user in the license. All invoices are first send to the bot’s worklist and what happens after is depending on how it is implemented. If an order is found and no mismatches are seen, the invoice does not have to be additionally approved. It is still possible to send such invoices to your coders to check and edit.
To be able to use this bot user, please let your application manager contact the service desk team.
Scan optimization requests
Onventis Invoice Processing leverages advanced OCR technology to simplify and automate invoice processing. This technology identifies data on PDF documents and stores it within the application, ensuring that invoice information is recognized upon receipt in the Worklist.
Create Optimization Requests: If the OCR technology incorrectly recognizes any data, you can provide feedback to correct it. In the Worklist of coders, click on the three horizontal dots in the upper right corner of the document and select "Optimize recognition" to make adjustments.
You will see a pop-up where you can provide feedback on the following fields to improve or complete recognition for future instances:
VAT Number
Chambers of Commerce number
Bank account number
Invoice Amount
Invoice number
Invoice date
Line recognition (if applicable)
After providing feedback on the field to be optimized, press "Send" to submit your request.
Feedback on optimization requests
To access historical optimization requests, users need the "OCR Optimization Viewer" role.
Optimization requests will be processed within a few working days. After feedback is provided, you can view it via the "OCR optimization requests" menu path, accessible from any invoice. A pop-up will display the status of previous requests and the response message. An example is provided below.
Pop-up with feedback: