To deliver video services that meet or exceed customer expectations, service providers must act quickly when quality and usage KPIs deviate from their normal range, IBC2019 exhibitor Agama Technologies believes.
Agama’s AI Anomaly Detection feature is said to automatically identify anomalies based on information from every subscriber and to provide actionable alerts, clear visualisation of detected anomalies and powerful interactive analytics. The feature employs automated self-learning to recognise patterns in video delivery networks.
Johan Görsjö, director of product management at Agama Technologies, said: “Separating actual anomalies from normal variations in KPIs is an excellent example of how AI and machine learning can be applied to video service assurance in a way that addresses real world needs.”
Acting on information collected in real-time from as many as several million client devices, such as set-top boxes and OTT player applications, the algorithm predicts how each subset of the population, from whole countries down to individual neighbourhoods, will behave based on past observations.
By detecting real anomalies and putting them into context, Agama says the solution creates situational awareness that enables faster analysis and problem resolution.
Agama is exhibiting at IBC2019 on Stand 5.B72.