Situation: Zixi is a global leader in the delivery of broadcast quality live video over IP. Their ZEN Master software platform is used by over 1000 media customers to source, manage and distribute live events and video channels over the internet.
Although video consumers demand the highest-quality viewing experiences, issues can occur during the transport of packets between video sources and target destinations. These issues can result in video quality degradation, video stream failures and other negative events. Zixi collects over nine billion telemetry statistics every day. So, extracting meaningful insights from data to get ahead of negative events is a huge challenge.
Solution: Zixi uses advanced analytics, machine learning and AI from nD to help quickly interpret vast quantities of streaming data to determine what needs attention. Users can quickly identify which customers, channels, video sources and targets are having health issues and drill in to determine root causes.
nD is also being used to create predictive models that alert users to the likelihood of certain events happening in the future, such as stream failure and content quality issues. Zixi also uses anomaly detection and smart alerting to alert users to deviations from normal behavior and help engineers to more quickly pinpoint errors and identify trends.
In addition, incident attribution models are used to quickly determine root causes of systemic issues and propose remedial actions. This includes Multi-Object Correlation Analysis (MOCA) that automatically analyzes incidents across all channel sources and target destinations. Incidents that share a common attribute are correlated, helping operators rapidly identify probable root causes and understand their impacts across seemingly unrelated channels. Find out more here.