faktur:digital saves 79% of its time on document processing for its clients with Rossum

The Challenge

faktur:digital GmbH is a German business service provider. It offers its clients solutions in the areas of document management, document workflow, input and output management, and archiving.   In its portfolio of compact and cost-effective business solutions, faktur:digital has a proprietary cloud software called faktur:box Archive for DMS functionality that closes the document processing lifecycle. However, the OCR component of their previous IDP solution for invoice recognition didn’t meet the expectation about the extraction quality and accuracy out of the box.

To find new, more reliable solutions, the team evaluated a total of three companies to provide image-to-text data extraction capabilities. Importantly, the possibility to seamlessly embed the solution into faktur:digital’s own systems were central for the team. Rossum became the service of choice for invoice data capture because of the best AI outcomes of the tests conducted and the flexibility of AI for use cases that the pre-trained AI engine couldn’t process.  Finally, Rossum was the only one out of all solutions evaluated that offered the embedded mode, which was very important for the team.

The Solution

At first, faktur:digital embedded Rossum’s pre-trained AI engine to cover the part of its workflow for invoice data recognition. To adjust the solution to fit the specific business requirements, the team initially leveraged serverless functions that allowed for faster implementation combined with high performance and good usability. As the customers started inquiring about other document types, the team at faktur:digital partnered with Rossum’s support team to implement dedicated AI support for additional document types such as delivery notes and order confirmations.

For faktur:digital’s clients, the input channel varies from invoice inboxes to scanning the paper invoices, with some customers using the drag and drop upload for their documents. No matter the channel used, the document is uploaded to a microservice which then manages an upload into Rossum, where the AI performs automatic checks and extracts the data. Due to faktur:box Archive tight integration with Rossum, the system would flag the need for human intervention. The customer then uses Rossum’s embedded mode function to perform data correction whenever human validation is required.

The Impact

The seamless automation allows faktur:digital to reduce its data extraction times by almost 80%. “It´s important for every company to use its employees’ time as effectively as possible. Binding people to manual data extraction and entry from tens of thousands of documents is a time waste. For us, AI automation means employees now have the time for more specialized tasks, which makes the business more efficient,” – says Alexander Berndt, Cloud Engineer at faktur:digital.

Additionally, speeding up the invoice processing allows faktur:digital’s customers to realize early payment discounts that sometimes amount to 3-5% (and more) of the total payment due.

faktur:digital plans to grow its business by 15-20% yearly in document processing volumes, and Rossum will play a big part in this growth. “At the moment, we are sure that every data extraction problem could be solved with Rossum.” –  says Alexander Berndt