Follow us in this blog and learn the latest about the functionalities of our new extraction plugin for Ephesoft.
We have developed this plug-in for Ephesoft Transact to massively optimize the extraction of metadata by using Artificial Intelligence and enhanced OCR technology.
The FellowKV² plugin solution is a cloud-based repository that gives customers access to a pool of extraction rules for header and footer fields. Through machine learning, the system is able to learn new extraction rules more effectively and autonomously each time a document is uploaded, validated and exported.
Input from all of the plugin’s customers continuously improves the system, as does ongoing training of the AI by Fellow Consulting AG. This means that every time a company trains the IDM Capture-Ephesoft software with a new extraction rule, that rule is made available to everyone in this cloud-based environment.
This process is performed in the validation window, where all companies have the possibility to correct/add/delete the extracted data from all available fields. Once validated and exported, Fellow Consulting will manage this new extraction rule defined for a specific document layout so that the rule pool is populated and updated for further use.
As the name of the FellowKV² plugin suggests, it is a plugin that needs to be installed for the extraction and export module in the IDM Capture – Ephesoft software.
Once installed, the solution can be used and trained from day one without the need to configure the extraction rules batch class anymore. We will continue this blog series with the technical details about the plugin and the AI used.
Learn even more about our new PlugIn here and watch an explanatory video about it!