Empowering Managed Service Providers: Decentralised AI-Enabled Monitoring in Multi-Tenant Networks
Adeel Rafiq, Muhammad Zeeshan Shakir, David Gray, Julie Inglis, Fraser Ferguson · 2024
Managing a multi-tenant network presents formidable challenges to managed service providers (MSPs) as they endeavour to provide exceptional service quality to clients scattered across diverse locations. These challenges are compounded by clients' stringent requirements to shield the identities of their network elements, preventing their exposure for centralised monitoring purposes. Consequently, the task of MSPs in overseeing geographically dispersed networks becomes more intricate due to the necessity of establishing separate real-time monitoring systems for each client. Moreover, the importance of visualising monitored data cannot be overstated, as it unveils invaluable patterns that illuminate service quality issues. Consequently, a monitoring tool equipped with a richly informative dashboard becomes indispensable for strategic planning and the delivery of top-tier services. Furthermore, the integration of AI -driven resource prediction enhances incident resolution capabilities by enabling proactive alert notifications. In response to these complex challenges, this research introduces an architectural solution centred around a data visualisation platform. This platform harnesses the power of a portable, decentralised AI-enabled real-time network monitoring tool custom-tailored for multi-tenant networks. The proposed approach adopts a pragmatic strategy that seamlessly integrates multiple open-source modules across geographically dispersed networks. The result is nothing short of remarkable compared with the manual approach, with network element downtime experiencing a remarkable 95% reduction, and incident resolution seeing a noteworthy 90 % reduction. This transformative impact directly benefits service desk professionals and network design engineers alike of MSPs.