SS Data stream

Goal

to provide data lake for BI tools with relevant data about players and their actions.

Background

We use SS functionality for Event Streaming, SS-Event-Streaming-Get_started.pdf

Technical decisions

Soft Swiss provides Kafka Stream with real-time feed of events from the platform. It's a big step comparing to previous (SS-data-export) implementation, where we have to work with data spited by part with a hour granularity.

Also now we receive not snapshot of data, but feed of profile changes. This allows to get actual profile data, or even more – we can get feed of group updated. For example,we can track optin/optout to different groups for our users. With this, we can check which users were in a high roller group in exact time. Because of big amount of data we use BQ as Data warehouse.

Data structure: