Data Centre ICT Compute Product of the Year
The Traffic & Service Delivery team of a leading Asian telco works with massive volumes of multi-sourced data, coming from network probes, radio as well as users' geolocation data.The mobile operator was interested in being able to geographically map its network Quality of Service (QoS) by cross-referencing network quality data with users' geolocation details. This would enable them to pinpoint which locations suffer from poor QoS. The task was infeasible with the company's existing infrastructure, which would run over ten hours trying to execute the complex JOIN operations. To be able to run the operation on a daily basis, the time to insight would have to be shortened significantly.
The SQream data analytics acceleration platform was implemented to accelerate the ETL and query processing. The multi-sourced data is continuously ingested directly into SQream at an accelerated rate of 3 TB/hr. At the end of each day, the Traffic & Service Delivery team runs the complex JOIN query on SQream, which successfully completes in under an hour – a 90% reduction from the ten hours it took before SQream's implementation. The transformed data on SQream is further fed to their visualization platform for end user consumption.
The team is now able to see the quality of service at any geographic location, and immediately pinpoint locations that are facing QoS issues – both at that moment and over a period of time. With this knowledge, relevant engineering teams are triggered to proactively fix these issues before customers calls to complain. The company's ability to proactively improve QoS has led to significant improvements in customer satisfaction and reduced subscriber churn.