AI is transforming the telecoms industry. But training AI engines requires huge volumes of data. Active Monitoring is part the solution, because it provides a constant source of realistic data from which models can learn.

AI is already transforming the telecoms sector and will continue to do so. The complexity of today’s networks and the services they carry simply require AI. One challenge is training AI engines. Active Monitoring is a key part of the solution.

There is no doubt that Artificial Intelligence will transform the telecoms industry – it already has. The deployment of increasingly complex network infrastructure and the dynamic, advanced, differentiated services (such as network slicing) that 5G brings means that the only way to manage and optimise the network, and the services it carries, is through a combination of Automation, AI, and Machine Learning.

AI content will account for most network traffic in the next few years

Technology consultancy Omdia predicts that by 2025 most network application traffic will involve AI content generation and processing, and by 2030 nearly two-thirds of network traffic will involve AI, mainly video and image content. It means that using AI themselves to optimise and manage the network is even more important.

As a result, an IBM Institute for Business Value survey found that most CSPs are assessing and deploying gen AI use cases across multiple business areas, including network management and optimisation, customer service, risk and compliance, data security, edge computing, to name just a few^[1].

Many are already using it. For example, a study by chip specialist Nvidia found that in 2023 90% ot telecom providers were already using some form of AI –  with 53% of them agreeing or strongly agreeing that adopting AI would provide a competitive advantage^[2].

When it comes to network use cases for AI, there are three main areas:

• Fault detection, prediction and resolution
• Network optimisation
• Network planning and upgrades

AI offers multiple benefits to network operators

But there are multiple other benefits AI has to offer. It can be used to predict where potential faults might occur in future, identify customer behaviour trends and even predict future trends, advanced data and analytics, the ability to offer highly personalised services, and optimised customer service, among others. A study from McKinsey, for example, suggests that AI can increase sales conversion by 15% and reduce capital expenditure by 10%^[3].

Put simply, AI can enable cost reductions by making the network highly efficient – through automated network management, eliminating network outages, and allocating resources to the right place at the right time – provide insight into network and customer behaviour, offer predictive analytics, identify potential future network outages, provide predictive maintenance information, supercharge customer service, monitor for potentially fraudulent activity, enable differentiated services such as network slices – the list goes on.

One of the main challenges of AI, however, is training it. AI needs huge volumes of data from which it can learn. Of course, it is relatively in its infancy still so training it accurately so that it can ensure all of the benefits outlined above can create a problem.

Active Monitoring can accelerate the training of AI engines

Emblasoft Evolver and its Active Monitoring capabilities can in fact be used to help train CSP AI models by injecting realistic user data into the network and observing its behaviour. This can be performed at huge scale, over multiple protocols, to mimic how the network responds under various user sessions, which AI can then learn from.

Put simply, data is AI’s fuel and Active Monitoring can be used to enhance your AI capabilities. By feeding the AI engines with realistic data, which is injected into the network, it’s possible to train AI models by learning from real-time events and live responses. Active Monitoring creates a positive feedback loop for AI engines, and it can be performed on a continuous basis. It’s another source of data that can be used to optimise AI models and discover new innovation opportunities.

Emblasoft Evolver’s Active Monitoring capability can enhance and accelerate AI programmes by injecting large volumes of realistic data into the network, from which AI engines can learn. To find out more, contact us today.

^[1] https://www.ibm.com/thought-leadership/institute-business-value/en-us/report/telecoms-generative-ai

^[2] https://resources.nvidia.com/en-us-ai-in-telco/state-of-ai-in-telco-2024-report?xs=582167#page=1

^[3] https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-network-is-the-product-how-ai-can-put-telco-customer-experience-in-focus