ICR for the automation of invoice processes

One of the departments in which most progress is being made in digitization processes is finance. Among its functions is that of accounting with the mandatory management of invoices. In many cases, these are processes that, even today, are done manually. To address how to improve the invoice management process, Byte TI organized a meeting with Yooz that included the participation of Verónica Maya, Yooz Advisor in process automation.

The meeting began with the analysis of the reality of the accounting department and also of the professional offices and consultancies. When accounting for invoices, there are many cases that, for different reasons, are done manually. This causes processes to slow down, which affects the productivity of the department. The problem is that, for this task, many companies have outdated technology, which makes one of the most inefficient processes that currently exist in organizations appear from the moment a purchase is made, the invoice is generated and paid. This whole journey can take up to 35 days, according to Verónica Maya, Yooz Advisor in process automation. In addition, as Maya explained, “the manual processes of managing an invoice can involve a cost that is between 9 and 20 euros, to which around 3.25 euros should be added for archiving that document.”

This whole situation means that a lot of time is wasted and that the finance and accounting departments have to multiply their efforts to carry out the management processes of each invoice. “This is a heavy, complex process and there is a lot of room for improvement. However, with the automation of invoices, the financial department can become a great productivity center of the company”

It does not work with an OCR

Although manual processes still exist, technology has long been incorporated into these departments. With the introduction of OCR technology, those processes have been greatly improved. But, as Maya stated, “OCR is a technology that only performs character recognition. It is a fundamental technology to digitize data and collect information, what happens is that in many departments, it has become obsolete”. And it is that OCR has been evolving rapidly and now it is just one more element of a much more innovative technology such as ICR or “Intelligent Character Recognition”.

The ICR, unlike the OCR, makes it possible not only to recognize the data, but also to understand the accounting content of each invoice and give way to the automation of each one of them. KPIs are more flexible and with it there is no need to correct data. In addition, it is a technology that is learning with Deep Learing, which means that, by using intelligent technology, it will allow digitizing processes and automating the digitization of invoices, making manual invoices disappear”, assured the Yooz consultant.

The costs

According to a study by PayStream, the incorporation of this technology means that it was found that the processing cycle of invoices to be paid is 45 days in those companies that are beginning to use this type of technology, gradually and to the extent possible. that the software is learning to manage all the processes (reception, filing, processing, uploading to the ERP and payment) that surround an invoice in just five days. In other words, by implementing a technology that incorporates AI and automation, the costs of the accounting and financial departments are reduced by up to 75%”.

The ICR, unlike the OCR, makes it possible not only to recognize the data, but also to understand the accounting content of the invoices

As Maya stated, “OCR has fallen short for the current needs of financial departments, while the incorporation of these new technologies increases productivity and reduces costs”

The proposal

And what is Yooz proposing? In this case, we are dealing with a 100% SaaS technology that uses technologies such as OCR together with Machine Learning and Deep Learning to help companies manage their invoices. It was born as a 100% cloud solution and since its inception it has been growing around 50% per year. It currently has 200,000 users and more than 100,000 companies that work with it. One of its advantages is that the solution is perfectly integrated into the most popular ERPs on the market and, furthermore, it is authorized by the Tax Agency. “Our goal is to put all this technology into place quickly by providing a solution that is economically cheap and very simple to use. And it is a solution for all types of companies. Among our clients, we have small and large companies from all kinds of sectors, in such a way that the solution adapts to each environment and each need”.

How does it work

The accounts payable automation software takes advantage of the OCR of scanned invoices to extract data from all types of documents, whether they are scanned invoices, photos regardless of whether they come from any type of source that the client has chosen. It then interprets the information, performs the accounting imputation and sends it to validation stages if necessary. Finally, the entry that is generated is exported to the ERP or to the accounting system that the financial department uses. In other words, Yooz’s solution automates data capture, invoice registration, payment validation and data export. All this without the need for anyone to touch anything, unless the software finds a problem that must be reviewed by a human. Yooz then interprets the information and extracts the relevant data so that this data is reviewed for validation. Once the documents are validated, they go directly to be exported to the ERP and over time, Yooz learns to fill in the missing data on the invoice.

The process for the user is really simple as Yooz takes advantage of artificial intelligence and intelligent automation to make the process of handling an invoice really simple.

Thanks to the Yooz solution, the most cumbersome tasks of the financial department are totally eliminated. Yooz has an omnichannel capture capability and adapts to the way each department works. What is sought is that the client feels as comfortable as possible, so the graphic interface or the information that can be traced from the invoices can be modified.

By applying artificial intelligence, the platform knows, for example, what is the usual invoice arrival method. Thus, if they usually arrive via e-mail, they are uploaded to the platform automatically and without the need for human intervention. In this way the invoices are ready from the first moment. Not only that, but it can also be done from different ways of capturing: even from a photo from the mobile.

The advantages are varied, since even if, for example, a PDF is included that has different invoices, Yooz processes it and discovers all the invoices that are in the document, in such a way that the user no longer has to cut each page to form different PDF documents.

The value of Yooz therefore resides in the analysis capacity that the tool possesses and that does not require any control by any user. For example, if an e-mail contains different attached invoices and other documents, users can teach the solution to only incorporate those documents that contain the word invoice into the platform and discard the rest. In the event that the tool does not correctly detect whether or not a document is an invoice, the tool will automatically take it to a specific folder so that a user can see if it really is an invoice and incorporate it into the accounting system. In other words, at that moment, the user is parameterizing an invoice and the next time the tool is faced with a similar case, it will already know what category that document has.