Speakers

With thousands of interactions per hour, and while normalizing player events and feeding them into different internal tools, how do you spot what is abnormal behavior?  And how do we define what is abnormal behavior? Tracking what matters, measuring it, and setting up alerts. And most important, keeping it private and secure, so that the player's information – even if just statistics – is internal read-only. Within a hybrid network, using Cloudwatch, Prometheus and Grafana, and with a bit of help from our dev-friendly JavaScript, we're going to show you how we keep an eye on one of our most important queuing systems - data streaming from our Discussions platform database to different internal consumers of different type. Read More...