The traditional process of creditor administration
Many organisations still handle their creditor administration in a traditional way. That means there are often multiple employees active in this department to appropriately categorize and log all incoming invoices. Subsequently, the payable invoices have to be approved by managers from the respective departments before they are effectively added to the payment process. When you’re dealing with hundreds of invoices per week, is easy to make a mistake – and any mistake can have far-reaching implications. Therefore, we have developed a solid alternative which helps to reduce the workload for these employees.
How Artificial Intelligence safely takes over the process
Looking back at the example of creditor administration, the added value of such a technology is very clear. By uploading at least 30 similar invoices to the software, it will recognize where it can find specific information and how this information is displayed. This ensures that customer details are quickly found and added to the CRM system’s respective fields. Because the software is also able to read the invoice number and the contents of the invoice, everything can be booked in automatically. This in itself already saves a significant amount of time and reduces the risk of errors.
How the software can find the right data in documents, as well as in other systems
The ability to read invoices is one of the easiest processes for a robot, because invoices are usually constructed in a similar way – regardless of their specific contents. However, it is not inconceivable to also have the software read other document types. As long as the required information can be recognized, it is possible to train the robot to look for these specific properties. It is also possible to connect the software to other systems in order to extract data.
Humans will always maintain a monitoring role
If the robot is well-trained, it can take over at least 90% of the process – fully automated. For the remaining 10%, humans will always be needed. Once more looking at the example or creditor administration: new invoices will have to be entered in order for the system to learn. Scanned invoices can sometimes have some ink or dirt on them, resulting in the AI not being entirely sure whether it is reading a ‘W’ or ‘VV’ – or the address is changed. The primary benefit of this AI solution is that after a couple of human-made corrections, it will understand what it has to do. This clearly shows us how the robot learns, giving you the confidence that your documents are processed correctly.
In the end, humans will always maintain a monitoring role. But by automating the majority of tasks, employees will have the time to focus on other activities.
Would you like to know more about how our AI software works and how it would function in your organization? Feel free to contact us and request a demo.