Guavus Alarm IQ

90% reduction in network alarms with 0% disruption to your business

Guavus Alarm IQ  harnesses the power of AI to eliminate alarm noise without requiring changes to Network Operations Centers (NOC) operator work flows.  Using advanced machine learning algorithms, alarms streams are analyzed and grouped accordingly to common attributes to dramatically reduce the total alarm count.  Then these alarms are prioritized based on the probability of them leading to a negative event and therefore only alarms that are predicted to cause incidents are categorized with a high priority.   This reduces the number of alarms that NOC staff must manage by 90%+ and lets them focus on real issues without wasting time chasing meaningless alarms.


Guavus Alarm IQ

Most CSP NOCs receive a minimum of one alarm per second (~86,400/day), however, only 2-3% of the alarms actually lead to true incidents or problems.  The rest are simply noise that can and should be ignored.  The problem is distinguishing the noise from the signal.  Standard alarm de-duplication and management tools classify 10-20% of the alarms as critical (up to 17,280/day), which is still too many to handle in one day.  The NOC teams have no choice but to ignore the majority of these – but how do they choose which ones should be ignored?

Guavus Alarm IQ uses machine learning and AI to understand which alarms are truly important and will actually impact customer experience.  It automatically takes in streams of alarms, classifies them and predicts which 2-3% of alarms will lead to incidents. Now NOC operators only need to focus on a much smaller subset of alarms and can confidently ignore the rest.  A true game-changer.

Features and Benefits

  • 90%+ reduction in alarms
  • Automatically dismiss alarms that are just adding noise
  • Prioritize alarms based on customer impact using advanced analytics
  • 84%+ reduction in Mean Time to Understand (MTTU) issues
  • Correlate & group alarms based on root issue to reduce alarm overload
  • Dramatic increase in discovery of truly impactful alarms
  • Automatic periodic model retraining and model adjustment for accurate incident prediction
  • No change in user interface. Integrates with standard OSS tools such as Netcool and Remedy

Guavus Alarm IQ leverages machine learning to intelligently reduce your alarms

Support for NOC use cases includes:

Guavus Alarm IQ

  • Alarm Correlation and Prioritization
  • 90% Reduction in Network Alarms
  • Auto-dismiss Noisy Alarms
  • Root-issue Identification