Klippa, the Dutch specialist in intelligent document processing, recently announced that it will continue under the Doxis name. Rocking Robots spoke with Chief AI Officer Yeelen Knegtering about the background to the acquisition, Doxis’ strategy and the value of AI agents.
Originally founded in the Netherlands, Klippa grew into a major player in intelligent document processing (IDP): the automatic recognition, extraction and processing of documents. Doxis’ proposition complemented that. The company has long been active as an enterprise content management (ECM) system: the place where documents are ultimately stored and managed for large companies worldwide.
“Doxis did not handle the first step,” Knegtering explains. “They stored and managed documents, but automatically ingesting and processing documents was our speciality.” For Doxis, Klippa was therefore a strategic acquisition. The result is a platform that covers the complete lifecycle of a document: receiving, processing, storing and generating new documents based on available data.
Initially, the brand names continued to exist side by side. But as more customers started using both products, that became increasingly difficult to explain. “Then you have to start explaining it,” says Knegtering. The decision was made to move to one name, one platform and one experience. Klippa therefore continued as Doxis.
Billions of documents
Doxis’ customers are organisations that process large volumes of documents: insurers, banks, logistics companies and government bodies. The challenges they face are often similar: the process is expensive, difficult to scale and hard to keep compliant.
“Documents move around the world 24 hours a day, but people only work eight hours,” says Knegtering. “That leads to backlogs, long processing times and a lack of control over who has access to which information. Doxis brings that entire process, from intake to registration to output, together in one platform. That way, you can maintain control even when dealing with millions or even billions of documents.”
An additional advantage is that, because documents live in Doxis but are accessible from SAP, Oracle or other ERP systems, organisations do not have to completely overhaul their existing IT landscape.
AI in practice
Asked where customers are already using AI in practice, Knegtering says: “We see a lot of invoice recognition and automatic processing into ERP systems. Applications in contract management are also interesting. Suppose you have one hundred thousand contracts in your archive and want to conduct an annual contract audit. The legal department is not going to read all those contracts. AI can make an initial selection. Which contracts fall within the standards, and which do not? People then add value where it becomes complex and make the final decision.”
According to Knegtering, two developments are on the roadmap for the coming year, both centred on AI agents. “The first is an agent that implements Doxis for you. Organisations no longer have to manually click through endless settings menus. Instead, you describe in ordinary language what kind of organisation you are and what matters to you, and the agent proposes a configuration. You can still interact with it, adjust it, and then it sets everything up. That gives you a fully configured account without a complex implementation.”
The second development is a library of ready-made agents for common processes. “Think of an invoice processing agent, a contract management agent or an employee onboarding agent. The only thing an organisation still has to do is indicate which systems it uses, such as an HR package or accounting software, and arrange the authentication. The rest has already been done. You start at 80 percent; you do not have to reinvent the wheel.”
Groningen
The AI lab is based in Groningen, in The Netherlands. The University of Groningen plays an interesting role in this. “It attracts talent from all over Europe. People come there to study because it is a well-regarded university, with an above-average number of IT students.” And because Groningen has less competition in AI than some other parts of the Netherlands, Doxis has an advantage there. “We do not have to relocate that talent from all over the world. They come to Groningen by themselves.”
Randstad
Many organisations internationally already use Doxis technology. Randstad Germany, with a workforce of 57,000 employees, depends entirely on constant access to personnel files. Its existing archive suffered from poor performance and high maintenance costs, while the architecture had become outdated. Doxis fully replaced the old system: 27 million existing documents were migrated, with tens of thousands to hundreds of thousands of new documents added every day. The system went live just two months after the contract was awarded. More than 2,500 employees now use it daily through an SAP HCM integration.
Deutsche Bahn
DB Systel, the IT subsidiary of Deutsche Bahn, faced a classic enterprise problem: dozens of DMS and archiving systems running in parallel across the group, without a uniform standard. The solution was extensive: Doxis was designated as the group-wide ECM standard, including archiving, document management, workflows and email processing. The result consists of more than 500 million archived documents and over 40,000 users within the DB Group. Because everything has been standardised, new DMS projects can now be implemented within weeks.
DHL Express
DHL Express wanted one uniform solution for document management worldwide, replacing a patchwork of decentralised systems that did not provide consistent access to electronic documents and offered limited scope for expansion. Together with Hewlett-Packard, DHL built one of the world’s largest commercial document archives based on Doxis. The archive now contains more than three billion documents, spread across 66 document types. Every month, 130 million documents are added. Employees, customs authorities and customers around the world have direct access to the relevant documents.
These cases illustrate the scale of the projects Knegtering and his colleagues work on: managing hundreds of thousands to billions of documents. The scale is large and the sectors are diverse, but the challenge is similar: maintaining control over information, accelerating processes and staying compliant. The technology for this is being developed in Groningen. Knegtering says: “Anyone who wants to maintain control over their organisation starts with the documents. And anyone who does that well lays the foundation for everything AI will bring next.”
