Michael Mattis

Extract problem-specific content from operational chats, convert it into structured workflows, and make the data usable.

More and more frequently, chat apps are spreading as a means of communication in companies. In rare cases, this happens with proprietary systems and end devices, but most of the time standard applications are used on partially private mobile phones: the chat is used as a quick, informal medium. In the process, company-relevant data not only ends up on third-party servers, but problem reports and the shared knowledge for solving them are lost, even though they were written down in the chat.

In TexPrax, text analysis methods are used to extract the relevant topics and problem-solving data from the chats. Not only are technical solutions shown, but also the boundary conditions for their responsible use are developed.