(the client did not want to be named)
The client’s initial situation
The company needed to separate is processes from those of the rest of the group, and this change had to be made in an extremely short period of time.
The project included the migration of a solution for processing supplier invoices from Kofax/ReadSoft. The migration included an upgrade of all of the considered components to the latest versions of Invoices and Process Director (over six versions of the software).
IXTENT designed the company’s migration and upgrade solutions. The solution consists of an:
- SAP part (Kofax ReadSoft Process Director), which aims to optimize the process of processing supplier invoices,
- OCR part (Kofax ReadSoft Invoices), which provides automatic recognition and data mining.
With this solution, employees of the company do not have to manually rewrite invoices to SAP, and they also do not have to carefully check the content of each invoice – the system runs automatic checks against SAP data, the approval process is electronic and has a fixed order. This automated solution reduces administrative process costs by 50 % or more.
The solution consists of the SAP part (Kofax ReadSoft Process Director), which aims to optimize the process of processing supplier invoices, and the OCR part (Kofax ReadSoft Invoices), which provides automatic recognition and data mining. With this solution, employees of the company do not have to manually rewrite invoices to SAP, they do not have to carefully check the content of each invoice – the system runs automatic checks against SAP data, the approval process is electronic and has a fixed order. This automated solution reduces the administrative process costs by 50 percent or more.
How does it work? Description of invoice extraction process …
All received tax documents (invoices) are accepted as PDF files as e-mail attachments. As soon as a new e-mail message appears in the mailbox, the attachment is extracted and processed by the extraction engine.
Once the data mining is complete, the data along with the invoice image are sent to the verification module. At the moment, the only activity that is expected from OCR users is the formal inspection of images of tax documents against the relevant extracted data in the Verify module. After the tax document is approved (or rejected), the data are transferred to the SAP system.
Data mining from the tax document image is performed in the background without user participation. The document is prepared in the Verify module within seconds after it has been properly processed by a job for retrieving e-mail attachments and extracted. The basis of the interpretation is the identification of suppliers and the related templates of tax documents. This identification is based on three types of data:
- supplier’s logo
- supplier’s company Reg. No. (domestic) or VAT No. (foreign),
- supplier’s bank account details.
If there is a match in one of the mentioned types of data between the data mined from the tax document image and the stored data in the tax document template for the relevant supplier, then the template is selected and the data for the given vendor are fully extracted (extraction of individual fields from the image of the tax document).
The self-learning system of this product contributes to its greater success.
If a tax voucher from a particular vendor is processed for the first time, a recognition template is created for it – the Verify module operator must give the voucher template a name. The right template selection is critical to the success of the extraction. Therefore, the templates are optimized in the Optimize module to maximize the success of vendor identification and data mining.
Benefits of the solution
The company bought a relatively expensive solution for “only” 10,000 invoices/in the lower tens of thousands of processed invoices per year; however, the investment paid off. The measurable return on investment in this case was one and a half years, which is equal to the savings on the labor costs of two accountants (saving of 50% of the costs of human resources).
Course of cooperation
The project took place without complications, and was successfully completed within the required accelerated deadline.
The project was completed in a very short period of time thanks to the rapid cooperation of both parties. IXTENT is one of the few suppliers to have the competence and experience with this outstanding technical solution.Štěpán Švec, Account Manager at IXTENT
The work on the project has enriched me with further experience in the area of licensing rights. Thanks to the project, I had the opportunity to meet the wonderful people with whom it was an honor to cooperate. I will look forward to any further cooperation.Hana Horálková, Project Manager at IXTENT