The A.I. Visualizer

Making Machine Learning Graspable

Client:

Arago

Location:

Frankfurt | Germany

Date:

May 2016

Developed for leading A.I. company Arago, this app allows for the easy visualization and exploration of complex data structures used in the automated maintenance and supervision of large-scale IT infrastructures – both for system specialists and for non-expert users.

Project Phases:

Strategy | Concept | Design | Execution

Role:

Data Visualization | Interface and UX Design | Prototyping | Software Development | System Integration

A visual front-end to Arago’s underlying HIRO system, the aim of the application is to make the system’s navigation as simple as possible. The intuitive interface presents crucial information and system changes in real time.

Arago’s A.I.-based business process automation technology handles services such as IT infrastructure, software, applications, business processes, workflows and transactions. Through the A.I. Visualizer app, all information is accessible in a hierarchical order. At one glance, it shows the dependencies and automation of each component, for example individual servers, their installed applications and the stored data.

The interface provides several views and widgets to monitor, explore and supervise the machine learning progress, allowing accurate interpretation and fast decision-making for teams and individuals of various backgrounds: from marketing and IT managers to system administrators and data scientists.

Each entity is represented with all dependencies in the central graph view. Ongoing processes and the status of individual components are displayed in real time.

Users can examine and modify the underlying decision-making logic for all automation components via an easy-to-use graphical editor.

The app helps users to understand the progress of the A.I.-based automation that has been applied to each component and which is represented in “Knowledge Tokens.”

For maximum flexibility, the software is designed to run on a range of different devices – from single-screen laptops and tablets to multi-screen setups in control room environments.