Franziska Lorenz—
UX Designer

Construction kit for AI applications

Redesigned ExB's internally used platform to make its text mining and cognitive computing applications accessible for customers

Project Type

Rework / New feature

The project

Since ExB's main product has developed extremely fast, it was time to make it customer-ready. The product is a generic NLP based machine learning / artificial intelligence platform called Cognitive Workbench (CWB). With it, ExB enables its customers organize, configure and modify cognitive computing applications to find new answers for difficult problems. The main benefits of this application is the computing power, the capability to capture large amounts of data and that it combining text, images, documents and structured data.

Until 2016, the CWB was an internal tool of the company. I was part of the ambitious project to redesign and prepare the platform for customer use. One challenge was that the system should continue to serve our research employees, and it was clear that it would evolve. At the same time, users without prior knowledge of artificial intelligence or machine learning should be able to use it.

The Cognitive Workbench 1.0 - an internally used research and development tool

Average reading time: 6 minutes (1009 words)


Configuration mode

Configuration mode - there are data inputs (sources), combined NLP algorithms for information extraction (Thinker) and Dataset blocks that can be freely linked with each other. The last right column is for sharing data with other people in the team.

Dashboard - Scenario Overview & Activity Stream

Dashboard: Overview of all Scenarios and the current activities

Navigation concept

Navigation concept: A type of breadcrumb menu allows the user either to move back to the next higher level or to jump between output methods, scenarios or dashboards at any time.