Anomaly Detection

Case Studies

Anomaly Detection

A Telco Case Study

Business Challenge

To efficiently  monitor and manage their data center’s critical resources and capacity in real-time – seamlessly scaling with no down-time


Technical Challenge

  • Processing 100s of thousands of messages from multiple streams in sub-second latency
  •  Triggering analytics based on anomalies
  • Identifying routers, switches, servers, storage and DB events with window based computation
  • Predictive analytics



  • Streaming data from multiple systems and sensors are  ingested in real time
  • Anomalies trigger advanced analytics models to run
  • Analytic results trigger business logic to automatically alert or impact business process



  • Improved network performance and reduced data center downtime
  • Reduced TCO –  maintenance and IT costs