The special feature of the solution is its text-based approach, thanks to which relevant information can also be extracted and classified from unstructured data. Customers benefit from a degree of automation of almost 100 percent as well as up to 80 percent cost and time savings in document processing.

Evy Solutions was founded in 2017 by Michael Vogel and Arian Storch. At this point, ten years of development time had already gone into the original product idea: an app based on the artificial intelligence developed by Arian Storch himself. "We wanted to create an app that brings together documents from the analogue and digital worlds. Whether letters or emails - everything bundled, classified and correctly assigned in one app," recalls founder and CEO Storch. "But then, instead of this complete B2C solution, individual parts of it received requests from the B2B sector."

So in 2018, the start-up split the app into its individual modules and developed them further with a focus on the B2B application area: Evy Xpact was born and went live in 2020. Evy Solutions' customer base now includes many well-known companies.

Unique selling point: analysis of unstructured data

There is already a wide range of document analysis tools on the market. "But apart from us, there is no other provider that can also perform the analysis with unstructured data and extract the relevant information regardless of the document structure," explains Michael Vogel, co-founder and Managing Director of Evy Solutions. "Our solution doesn't care where the relevant information is located in the text," Vogel continues.

This so-called text-based approach is made possible by the self-developed AI, which has a natural understanding of language and identifies contextual relationships instead of stereotypical text positions and constantly optimizes its own speech recognition through self-learning. As a result, the software achieves 99% reliability after just three to four weeks of use (and therefore more than any employee), reducing the manual processing effort to one percent of documents.

Quick to deploy and 100 percent GDPR-compliant

The solution scores with its high learning speed. For example, whereas a conventional AI requires around 100,000 documents to "learn" a corresponding customer case, Evy Solutions only needs 50 to 100 documents for the same case. This ensures that customers are able to work fully with Evy Xpact very quickly - within four to six weeks. The self-developed AI "Made in Germany" also guarantees complete GDPR compliance. All data is hosted and processed within Germany.

Another major advantage is that Evy Xpact can process a wide range of document types and formats. These include transport orders, delivery bills, purchase orders, tax assessments, contracts, damage reports and certificates. Depending on the application, the product can also be flexibly configured from the individual modules: From classifying, sorting and filing documents to separating the content of batch documents, extracting information or checking the extracted content and, if necessary, enriching it with missing information through to semantic searches for content and correlations, Evy Xpact covers all areas in a modular system.

Example transport orders

In the first step of the workflow, Xpact Logistics automatically monitors incoming documents (e.g. email inbox), classifies the incoming documents and filters out the transport orders. All other documents are forwarded to other departments if required or remain untouched by the software. The software then reads the relevant information from the transport order (e.g. customer number, weight, size, etc.) and checks it for accuracy and completeness if required.

Evy Xpact compares the data with the company's system, validates it and enriches it with master data if necessary. If all the information is correct and complete, it is stored in the company's existing ERP/TMS system and the responsible employees are informed that processing has been successful or the transport order is executed. If the information is not correct, the software automatically sends it to the responsible person for manual revision.