To support an increasingly digital-first audience, our client, UK’s leading telecom provider, decided to ramp up its technology capabilities. Aimed at delivering higher customer satisfaction, the client sought to implement a modern solution to an age-old problem – repeat truck rolls and a high number of ‘Dead on Arrivals’ and ‘Early Life Failures.’
The existing setup consisted of multiple service management and network monitoring platforms, where over 95% of the data lay unutilized) to drive proactive diagnostics. This resulted in increased customer calls, repeat truck rolls, a high number of Dead on Arrivals and Early Life Failures. Key challenges included:
- 20%+ repeat truck rolls due to repeat (continuous) faults and lack of fault correlation for service degradation
- 15% early life failure and DOA resulting in bad customer experience
- 65%+ manual intervention for fault resolution leading to missing SLA and need for human resources
The SolutionWe studied the current process to design a network fault prediction platform that met the client's need for flexibility and agility.
The platform included an AI-enabled self-learning intelligent algorithm that continuously carried out predictive analytics on data consolidated from different systems for proactive troubleshooting. It then monitored precursors of failures, such as traffic flow, capacity, and utilization in near real-time. Key features of the solution included:
- 15 AI models to predict faults and network anomalies across a wide variety of network types and devices
- Integration with 15+ service provisioning, assurance, and monitoring platforms
- 360-degree network topology view overlaid with business, customer & operational impact visualizations
- Predictive models to identify the main fault location, root cause analysis, correlation, and impact assessment
- Self-learning knowledge hub to automate next best action
- Modular platform to solve specific challenges across Network faults, performance issues, and outages
Virtusa’s network fault prediction platform facilitated business agility, scalability, and optimized performance.
The client realized the following benefits:
- 55% reduction in manual incident creation; reduced customer inbound calls.
- Reduced truck rolls with an improved association with proactive incidents.
- 65% improvement in DOA test during provisioning by using ML-based diagnostics improved customer experience achieving business KPIs of DOA reduction