Using real-time analytics in the Data Lake, we can now detect performance impacts in our CROSS 2 retail system before they get to affect users.
In the past, the world of support looked like this: a user would report that an action in the system was not running as it should do. Several teams would then begin an elaborate analysis – finding the final solution to the issue was often a time-consuming process.
An interdisciplinary team set itself the task of improving this situation. Their goal was to pro-actively detect restrictions without having to rely on a report coming in from users. The first step was to record in real-time important measured values such as the duration required for opening an order or printing invoices.
Visualising the Data Lake: such a complex picture
But this data is complex and difficult to visualise. The breakthrough here was achieved by means of intelligent analysis in the Data Lake. Log files are now continuously sent from CROSS 2 to our central data platform, where current measurements are compared with historical ones. This data is visualised directly in our support system, and if there are deviations, an alarm is triggered. In this way, support colleagues can see at a glance where performance restrictions are emerging, allowing them to intervene in the system immediately and pro-actively.
With this data-driven support model, our colleagues have raised the service quality for our 40,000 users in all 19 markets to a whole new level. A real milestone! The solution was so groundbreaking that the data-driven support project was honoured in our in-house competition. Congratulations to the project team for the Change Award 2020!