Exceptions to document review will not go unnoticed

Data Verification and Validation play a crucial part in the ongoing Digital Transformation that every organization deals with. It’s not an issue as long as the automated process meets the initial design requirements.

This article talks about data errors found in processed documents in situations, when the required data quality is not achievable at once. E.g., a trained Machine Learning model needs to be retrained to produce the expected results. In those situations, there is a need for an environment that provides practical solutions for Verification and Validation needs without further ado.

EasyVerify is created precisely for this, an intuitive and flexible web browser-based platform that supplies both experienced operators and the occasional users with tools to carry out document Verification and Validation.

Validation stands for data checking. If automatic validation fails, the user (operator) comes to make the judgement and corrections if needed. EasyVerify has all the necessary inventory for that. Our technology offers predefined business rules applied to the extracted data and the user can quickly make the right decision on accepting or correcting the data.

This processing step yields more than just document validation. While the operators validate the problematic documents, our validation technology takes notes behind their shoulders. Nothing the operator has to worry about. More of a technological advantage that the operator enjoys! Using manual fixes of the data issues that automatic validation stumbled over, the system learns and next time less manual work is required. The operator trains the setup for further optimization!

Error-free automatic data processing, that’s what we all want. We do our best to get as close as possible to this. In some cases, you see that we read the data correctly, think of handwriting recognition (ICR). We read a date filled in by hand and, with the accompanying documents, we come to the conclusion that it concerns a certain date. In this case, we do not read the date completely correctly, our ICR is clever, but it cannot handle all the scribbles. In such a case, the technology weighs the recognition at 70%.

So, we can automatically determine the certainty of 70% based on recognition and other accompanying documents. In this case, 80% certainty has been agreed with the client. In practice, this will mean that this document with the uncertain date readout is offered to an operator. Verification and Validation in optima forma!

EasyData mainly supplies Data Capture technology that is partly driven by Machine Learning technology. There are many models and datasets available for this. Think for example of Invoice Recognition. Over the years, a standard has been developed that suits your process! In addition, we know immediately that every company receives invoices and then processes them in a company-specific way. That is where Document Verification and Validation comes in.

Suppose you want to split a collective invoice per project or administration. Or even want to invoice directly. These are the steps that process automation makes possible with EasyData.If you like to be in control at all times, EasyVerify is suitable for that, data validation is at your service!

In many business processes, data not only needs to be verified, but values ​​also need to be checked. Take, for example, incoming forms with dates, numeric values ​​that appear on different pages and license plate information. Before the data kind of data goes to external systems, all has to be checked and validated.

EasyData software performs the required math calculus automatically. Our solutions are specially designed for this. However, some forms still need visual validation in case of exceptions, possible incorrect input by the user or possibly misinterpreted handwritten text.

With EasyVerify you can check and execute field calculations visibly. The architecture of EasyVerify allows you to execute complex mathematical formulas controlled by the operator’s eyes.

More information or a demonstration? We would love to hear from you!