Traditionally, managing the IP Multimedia Subsystem (IMS) has been a major challenge for Communication Service Providers (CSPs), as it has often been seen as an overly complex architectural framework with limited added value compared to the complexity of operating it. CSPs have always faced challenges with manual, tedious operations, and network management, as well as the time-consuming nature of fault detection and resolution. Combined with the significant infrastructure footprint of voice workloads, this keeps the total cost of ownership (TCO) high, exacerbating concerns about declining voice revenues, especially with the advent of Voice over LTE (VoLTE) and Voice over NR (VoNR) services.
At ng-voice, we are reversing this trend by changing the way CSPs manage their voice network infrastructure. Working with Amazon Web Services (AWS) and TELUS, we have demonstrated how to make network operations faster, more efficient and with minimal manual intervention.
The answer lies at the intersection of ng-voice's truly cloud-native IMS, built for the cloud, and generative AI. The demo of this innovative solution focused on on-demand observability, real-time alerting of fault occurrences and closed-loop operations, which are crucial for maintaining VoLTE services.
Building on the flexibility of ng-voice's Hyperscale IMS and its cloud-native architecture, generative AI unlocks a new level of network management for TELUS. The solution architecture shows how a cloud-native IMS combined with cloud-native AI enables the closed loop. There are two types of streams: on-demand observability, which allows understanding what is happening on the network, and closed-loop operations, which enables the activation of recovery and mitigation plans on the network.
On-demand observability means that users can submit natural language queries, which are then transformed by the Large Language Models (LLM) into specific SQL queries and presented to the user in a readable format.
The second part of the demo showed how AI triggers instant alerts when issues arise within the network, enabling faster resolution. The chatbot interface enables natural language conversations to diagnose issues, ask questions, and even initiate corrective actions.
For example, if TELUS experienced a network connectivity issue, it would traditionally require a manual investigation that could potentially take hours to identify and resolve. With this new system, generative AI can analyse the situation, summarize the problem, suggest solutions, and even implement them, all in a conversational format.
ng-voice's cloud-native IMS solution, empowered by generative AI, enables TELUS to achieve significant cost reductions by streamlining troubleshooting processes and automating actions that minimize downtime and optimize resource allocation. AI-driven detection and resolution of network issues, coupled with real-time insights and reporting via natural language interactions, ensures faster problem resolution, reduced outages, and improved network management efficiencies for CSPs. This evolving technology has immense potential to revolutionize the future of IMS architecture and network management, particularly for VoLTE services. At ng-voice, we are committed to pioneering the beginnings of AI in cloud-native IMS and telco environments.
Generative AI enables natural language interaction with the network, turning queries into precise actions and automating responses to issues. This allows CSPs to detect, understand, and resolve faults much faster than with manual processes.
On-demand observability lets users ask natural language questions about the network and instantly get structured insights. This means operators no longer need to rely on complex dashboards or SQL expertise to understand network health.
Closed-loop operations mean that once a fault is detected, the system can automatically suggest or execute recovery actions. This reduces downtime, lowers operational costs, and keeps VoLTE services running reliably.
The chatbot provides a simple way for teams to interact with the IMS using everyday language. It not only explains the issue but can also propose or even initiate corrective actions, reducing reliance on specialized teams.
TELUS was able to test how AI-driven closed-loop operations reduce downtime, streamline troubleshooting, and provide real-time insights. This showed the value of pairing cloud-native IMS with generative AI for future-ready operations.
