Using SAP HANA (a homegrown in-memory technology) SAP along with eBay developed a pattern recognition system to identify trends and anomalies called ‘signals’ constituting an ‘early detection system’ which automatically processes thousands of time series metrics used to identity usage trends at eBay.  The Solution was both delivering reliable technology as well as a delightful user experience. The main emphasis for design was to distill complexity inherent to the domain and bring the focus back to the needs of the users. The design challenges were to enable the business community within an enterprise to contribute in analyzing complex data and help identify anomalies and patterns.  This information had to be presented in a way that was engaging, fun, inviting, rewarding, and simple to understand yet reliable enough to get the job done.

 

The project kicked-off with a 2 day Design Thinking workshop with users and stakeholders from eBay and SAP at the beginning of project. That enabled us to get a sense of users aspirations and inspirations for design direction. During the synthesis phase  we started to think about what causes analysts’ working processes to become tardy and troubled. After understanding, we focused on how to simplify and accelerate their process. We realized a large part of the problem was to enable the user to see data on different using different data visualizations across multiple dimensions and navigate, all while keeping it simple and intuitive. We created a dial navigation along with filtering capabilities, users can quickly find what they want to see. It helps the user to navigate deeper into a hierarchy without loosing context. We provided multiple ways of visualizing data to find more a accurate method.  Also, that allows the user to immediately identify and define the signals with color-coding of different statuses. After analysts define pertinent signals, all signals are shown as historical actions on a single page to give a comprehensive understanding of the signals enable pattern finding.  Peers can contribute with recommendations to help diagnose problems and bring transparency, resulting in more accurate decision making.

 

As the online retail industry continuously grows, companies need to track large amount of transactions while maintaining market edges.