Press Release announces six major trends for intelligent document processing in 2022, a provider of Intelligent Document Processing (IDP) software for analyzing and automating document flows based on AI technology, sees six major trends dominating intelligent document processing this year:

1) Use of pre-trained models
This year, a growing number of suppliers will use pre-trained machine learning models for intelligent document processing. The use of pre-trained models can shorten implementations of IDP solutions significantly if the data used for training has many similarities with the customer situation. IDP solutions are already popular in the US and US suppliers are now eager to sell their solutions in Europe as well.

Pre-trained models for vertical markets offer new possibilities to achieve a faster ROI with Intelligent Document Processing. However, many models are mainly aimed at the American market and the documents used there. These models will not be very useful in Europe with different business processes, more variation in document types and other languages.

Richard Smit founder and CTO at “Simply copying American models for the European market will certainly not be sufficient for most European use cases. Retraining a local model for a specific use case may still require a lot of effort depending on the amount of training data required. Our experience shows that combining ML models with logic in the form of business rules, patterns and scripting often gives better results with less training data and in less time.”

 2) More pre-built use cases
This year, customers can expect even more pre-built use cases. A complete set-up will be available for some use cases, which shortens the lead time considerably. Pre-trained ML models will be very useful there. However, it will be essential that additional training of pre-trained models is possible so that the results can be improved if necessary.

3) Transfer learning on the rise
To be able to deploy IDP solutions faster transfer learning will be used more often. An existing Machine Learning (ML) model is used to train a new model, supplemented with additional training data to ultimately achieve better results.

In 2022, companies will also be able to develop ML models faster and more efficiently through cross-training. Industry knowledge and data about this industry is collected and shared by various companies. ML models can then be developed with a joint effort and multiple companies can benefit from an ML model for a specific industry. Companies can then adapt these models to specific customer needs.

4) Valuable human expertise better supported by IDP solutions
Process experts have in-depth knowledge of document-related processes and the type of documents used in their organization. To better utilize their knowledge, review and feedback options will increasingly be implemented in IDP solutions. As a result, experts are enabled to verify the automatic classifications and extracted data and make corrections or confirm the results. It is important that the solution offers a user-friendly user interface, tailored to the type of user so that it can also be used by employees without a technical background.

5) Growing importance of protecting sensitive (personal) data
Functionality for data protection in IDP solutions will become more important in 2022. As digital transformations take place in a growing number of industries, more and more documents are available digitally. This makes them easier to store and share but at the same time the need to provide functionality for properly protecting the privacy-sensitive data in these digital documents will increase as well in 2022.

Richard Smit founder and CTO at “Protecting sensitive data starts with functionality for detecting it in documents. It is important that these data can be anonymized or even pseudonymised, depending on the role or position of the employee who consults these data. The final step is to create an anonymized version for the distribution of a document. has had this functionality for quite some time because we have deployed it early on.”

6) Shift to new payment models
In the past, companies often paid for document processing software based on estimated server capacity. In 2022, these payment models will change. Pay-as-you-go is becoming more popular. Companies can then pay per page that is processed by the software. This lowers the entry barrier for medium-sized companies and also makes it easier for them to allocate costs per department.

Richard Smit, founder and CTO at “Partly due to the pandemic, we saw a growing number of companies that were working on digital transformation in 2021. The intelligent processing of unstructured documents is often part of this process. This trend was already taking shape at many financial institutions, but we are now seeing this interest extending to other branches as well such as healthcare, pharmaceuticals, logistics and government. Therefore, we expect the demand for intelligent solutions for document processing to increase further this year.”

About is an Intelligent Document Processing (IDP) solution for analyzing and automated processing of document flows based on AI technology. ensures greater productivity and a better customer experience by automating document flows within customer-related processes. integrates seamlessly with business applications and processes in the existing IT landscape, and can therefore process unstructured information in existing systems. With this, the company is responding to the explosive growth of information and the actions required to process data correctly, in accordance with privacy legislation (GDPR).

Like this? Share it with everyone: