A Data-driven Platform to Understand Customer Needs and Improve Services

A telecommunications company in the Maldives worked with NCINGA to introduce a platform that powers data collection, introduction of innovative services, and continued relevance in an increasingly competitive industry.

Industry: Telecommunications
Location: APAC

Introducing the Customer: Dhiraagu

Dhiraagu is a leading telecommunications company in the Maldives. The company offers a full range of communication and broadband services. Dhiraagu has over half a million customers. Throughout its corporate history spanning three decades, the company has established a reputation as a market leader introducing new products - especially in emerging digital areas. Dhiraagu’s vision is to inspire and empower their customers to “take on tomorrow” and thrive in the digital future.

Lack of Business Intelligence for Providing Better Quality Services

The biggest hindrance that Dhiraagu faced to its vision of providing superior quality services for customers is a legacy network and policy enrolment platform. These were creating several problems: Internet traffic is the foundation of obtaining business intelligence and service monetization. Dhiraagu lacked a more informed understanding of customer usage patterns to enhance the quality of their network, provide innovative service plans, comply with regulatory requirements, and defend the network from fraudulent actors. And as with any telecommunications company, the onset of the COVID-19 pandemic tested Dhiraagu’s network platform capabilities. An increase in network traffic and a higher dependency on bandwidth required an intelligence-based solution.

Thus began Dhiraagu’s collaboration with NCINGA. Dhiraagu decided to re-architect its network and policy enforcement platform entirely.

Goals

Dhiraagu and NCINGA identified a list of goals that the upgraded network platform must achieve:

Technology Implementation

NCINGA with Sandvine used a virtualized application deployment on low cost, highly reliable commercially off-the-self (COTS) hardware. This enables the policy enforcement platform to become hardware agnostic and achieve greater scalability. The Sandvine policy enforcement and deep packet inspection application uses supervised machine learning models that have been pre-trained in-house and validated for accuracy. These proprietary models and techniques are built upon 150 different flow parameters.

Employing these techniques, Sandvine is able to broadly classify encrypted traffic into categories (such as web browsing, video streaming, VoIP, etc.) and accurately classify unique applications within categories (e.g., Facebook vs. Instagram, WhatsApp vs. Lime, Netflix vs. YouTube, etc.) – even when the traffic is encrypted and ESNI is in use.

You can also apply machine learning techniques to distinguish between authentic traffic and traffic masquerading as something else. Crucially, these techniques do not rely on reading or extracting fields from flows. As such, Sandvine’s machine learning is impervious to encryption. Because many services are already encrypted and VPN use is widespread, these techniques will become essential – and will perhaps even account for the majority of traffic recognition – in the future.

A Platform That Provides Improved Analytics Capabilities, Security, and Customer Satisfaction

With the NCINGA-designed PCEF/DPI and PCRF platform in place, Dhiraagu has been able to overcome their business obstacles. They are now able to: Dhiraagu now looks forward to staying abreast with the competition in the Maldives’ fast evolving broadband services industry.