Together with Sanoma we initially created a video analytics platform. It can do advanced analytics of customer behavior from streaming data on Nelonen’s Ruutu.fi service. The current solution is being developed relentlessly and acts as a benchmark for future services to improve customer retention and conversion.

The prototype was made with Apache Spark for processing real time data, Apache Kudu and Impala to store and provide OLAP-storage for the processed stream of data and Superset for visualising the results. The described pipeline was implemented on top of Amazon Web Services using EC2 virtual machines.

“Due to our current analytics solution being discontinued, we needed to act quickly. The proof of concept designed and implemented by Emblica helped us select the right direction with reasonable cost and time, and it formed the base of our upcoming solution.”

Antti Pyysalo, Director, Online Development