Case Study: Rubicon Project achieves automated real-time business incident detection with Anodot

Rubicon Project Automates Real Time Business Incident Detection with Anodot

Rubicon Project, one of the world’s largest ad exchanges, processes massive real-time traffic—about 13 trillion bid requests per month across seven global data centers and 55,000 CPUs. Monitoring this scale and the complex, time-sensitive behavior of buyers and demand-side platforms exceeded what human teams and threshold-based tools could reliably manage, creating risk of unnoticed disruptions that could harm revenue and partner relationships.

Rubicon integrated its existing Graphite feeds into Anodot to gain automated, real-time business incident detection that learns normal behavior and seasonality, correlates related metrics, and surfaces actionable alerts. The solution was simple to deploy and use, replaced manual thresholds, scaled to tens of thousands of metrics and dozens of users, and enabled faster root-cause identification (e.g., linking revenue drops to blocks, timeouts or bid responses), delivering improved uptime and partner service while avoiding the months of internal development work.

Rubicon Project

Rich Galan

Director Analytics