Onventis Invoice Processing 2025.2

Release 2025.3 is planned for May 17, 2025 (Beta) and May 31, 2025 (Production).

Introduction: Streamline your invoice processing with AI Powered invoice coding

This release focusses mainly on AI-powered invoice coding. Rather than emphasizing the service itself, which has already delivered impressive results, we have enabled all our customers to integrate AI-powered invoice coding with their existing services seamlessly.

A sophisticated algorithm evaluates each invoice—whether it's a PDF, e-invoice (XML or hybrid) or an invoice via Peppol—to automatically determine the optimal processing route.

Additionally, even if you have set up column recognition or specific recognitions within the currently used document recognition service (OCR), AI-powered invoice coding can be implemented on top. This enhancement empowers all our customers to configure AI-powered invoice coding within their existing workflows, ultimately enabling faster invoice coding and automation of incoming invoices.

AI Powered coding: combined with existing document recognition service (OCR)

It is possible to combine our existing document recognition service with AI-powered coding proposals, offering the best of both worlds. By using AI-powered coding proposals, you can assure that all invoices always get a coding proposal. Invoices related to a PO are coded based on the PO, while non-PO invoices utilize AI powered invoice coding proposals. This allows you to automate the invoice processing even further. This is a big step forward in AP Automation!

🚀 Do you also want to start with AI powered invoice coding? Subscribe here and an account manager will contact you to upgrade your license and plan the configuration.

📺 Learn more about AI Invoice coding in this webinar.

AI Powered coding: Improved VAT code prediction

We have significantly increased the accuracy of our AI model's VAT code prediction. This improvement allows customers to automate the coding process for even more invoices, enhancing efficiency and reducing manual effort.

AI Powered coding: Grouped proposals

Invoices often contain numerous invoice lines, such as one for each product ordered. In many cases, these lines share the same financial dimensions. To create a cleaner overview, some financial controllers prefer to group these lines. Our AI coding engine now supports grouping multiple invoice lines into a single line, provided they share the same financial dimensions and tax segment. The basis on which these lines can be grouped, can be configured based on your requirements.

To utilize this feature, your consultant needs to make specific configuration changes.

AI Powered coding: icon now disappears after changing the value

When using AI-powered invoice coding, a proposal (e.g., for the general ledger) is indicated with an AI icon. Previously, the AI icon remained even after the user overwrote the AI suggestion. Now, the AI icon disappears when the user overwrites the AI suggestion, providing clearer feedback and reducing confusion.

Overview of OCR optimization requests

Users could create OCR optimization requests (to optimize the document recognition) within the application, but there was no way to view previously created requests, leading to uncertainty and duplicate requests. An overview of created OCR optimization requests is now accessible via the user interface. Users can access this overview from the invoice screen by navigating to the three dots in the top right corner and clicking on the “All requests” button. To enable this functionality, a new role called OCR Optimization Viewer has been introduced. This role is automatically assigned to all users who currently have permission to initiate a purchase invoice.

Feedback on OCR optimization requests

Previously, users could create OCR optimization requests, but they couldn't view the responses within the application. This often led to uncertainty about whether the issue was resolved or if further action was needed. With the latest update, responses to OCR optimization requests are now retrieved and stored within the application. Users can easily view these responses to determine if their issue has been addressed or if additional information is required.

Support for Gmail integration with OAuth

With Google's basic authentication for Gmail integration reaching end-of-life, we have enhanced Onventis Invoice Processing to support OAuth for Gmail integration. This security framework allows you to integrate your Gmail account without revealing your password, ensuring improved security standards.

Mobile app | Redirect to worklist instead of archive

We have resolved an issue in the Onventis Workflows mobile app where users were being redirected to the archive instead of the worklist in specific scenarios. With this fix, users will now be correctly redirected to the worklist. No app update is required.

Twinfield interface: adhering to API limits

Twinfield has introduced API limits, restricting the number of requests that can be sent within a specific time frame. Exceeding this limit would cause remaining calls to fail. A mechanism has been introduced to handle API limits. If the API limit is reached, the system will wait for a specified time before sending new requests, ensuring that all calls are processed successfully.

Sorting of various dimensions

The lookup of various dimensions, such as projects, was previously based on an internal ID rather than the user-visible ID. This has now been updated so that the sorting of dimensions is based on the dimension code (e.g., project code), providing a more logical overview of the results and making it easier for users to select the correct value.