Unlocking the GenAI Opportunity: 3 Key Operator Challenges and How to Overcome Them

April 2, 2024

Generative AI (GenAI) is growing rapidly among mobile network operators, and numerous vendors offer GenAI-powered applications and solutions.

Most of these are based on the concept of the co-pilot and target business users, such as in sales or support.

This co-pilot is an assistant accessed through a dialog box on the relevant application, typically a BSS solution. Users engage, even collaborate, with the co-pilot on tasks and challenges they seek to resolve. Based on the data on which the GenAI model was trained, the co-pilot provides support with insights and recommendations.

Operators have been anticipating the arrival of co-pilots who can also help them manage the network, and we’ll be taking a closer look at that shortly.

Three formidable challenges

For co-pilots to do their ‘magic’ well and right, they must be trained on and be able to gain real-time access to data that is:

  • Sometimes on-prem, sometimes in the cloud
  • Correlated for context
  • Trusted to be complete and accurate

But these are three tremendous data challenges for operators.

Cloud readiness

The platform processing the data to train the GenAI model needs to be cloud-native since some of the operator’s data is on-premises, and some are in the cloud.

If the platform isn’t cloud-native, extensive capture and processing resources would be required, making for a very expensive, often cost-prohibitive initiative.

Moreover, it will be tough to scale to the enormous volumes of data required to train GenAI models effectively.

Data correlation

Data needs context to have meaning for the GenAI model, meaning it needs to be correlated for training.

This is more important given the masses of data that service assurance captures, which is uniquely and profoundly valuable for telco GenAI applications.

However, since the operator’s data is typically dispersed across multiple, siloed sources, correlation is a cumbersome, time-consuming, and complex task.

Trusted data

The importance of trusted data must be recognized. Data that needs to be completed and accurate, whether from service assurance, the network, or other sources, will result in GenAI applications that provide unreliable and inaccurate outputs.

However, ensuring data trust can be challenging since the relevant systems do not always take the right actions to complete and verify the data.

How RADCOM overcomes the challenges

At RADCOM, we recognized GenAI’s potential to bring value to the operator’s business users and network teams early on.

Moreover, the above three data challenges could potentially prevent all these stakeholders from taking advantage of the opportunity.

The good news is that we can help. Our automated assurance platform is inherently designed to overcome these challenges – it is cloud-native; it correlates all the masses of data that it captures; and, because the platform takes advanced actions to ensure data accuracy, such as digesting various inputs and performing smart sampling, among others, the data is wholly trusted.

Bringing the GenAI promise to life

With a cloud-native platform and correlated, trusted data, we have been focusing on developing high-value GenAI-powered co-pilot applications for operator stakeholders, including their network engineers.

We call these co-pilot apps RADCOM NetTalk, as they enable users to talk with the network to gain ongoing strategic insights.

RADCOM NetTalk for quality monitoring and network health

Incidents are an all-too-frequent phenomenon for network operators, and they can undermine customer satisfaction and cause financial and reputational damage.

The NetTalk co-pilot enables them to mitigate the risk, both during the incident and even before it occurs.

This co-pilot mines all the data captured by the RADCOM automated assurance platform and correlates it to data in the operator’s BI systems. It narrates the story to engineers, creating and explaining graphs and charts, delivering insights, and providing recommendations.

This way, engineers can identify the root cause of a network issue, develop the right remedy, and determine whether this is a one-time incident or a recurring failure faster and more accurately than ever before.

In addition, network managers and executives can also engage with NetTalk to receive insights about the potential ramifications of a technical issue, optimizing business and operational decisions.

For example, they can receive predictive inputs on how a rollout acceleration will likely impact costs and customer experience KPIs in a specific market.

RADCOM NetTalk for standards adherence

This co-pilot mines publicly available standards and other relevant documents and releases, such as 3GPP release updates, and empowers users with relevant insights quickly and intuitively.

This NetTalk app will also be able to correlate standards-related data with network issue-related data aggregated by RADCOM automated assurance to inform engineers whether a failure may be related to a configuration or setup that doesn’t adhere to industry standards.

With this capability, extracting the required insights from all these highly voluminous standard sources quickly and efficiently is possible.

RADCOM NetTalk for churn prediction  

The RADCOM automated assurance platform has access to data that includes every subscriber’s complete profile and every step in their engagement journey with the operator.

The NetTalk co-pilot for churn prediction mines the subscriber’s technical KPI data to identify behaviors of interest and correlates this data with the subscriber’s responses to surveys, contact center call summaries, and churn lists.

Then, a satisfaction and churn risk model is created to reveal sentiment that has yet to be expressed but indicates the likelihood of churn.


These are exciting times for operators with the advent of tectonic shifting technologies and the groundbreaking applications that leverage them.

Generative AI has the potential to revolutionize almost every aspect of business—from marketing and sales to care and support and now network management as well.

The revolution is already underway. With the RADCOM quality insights co-pilot and standards adherence co-pilot, network engineers can access previously unattainable insights and recommendations for assuring network performance and do so in real time. With our churn prediction co-pilot, the business is positioned better than ever to protect its customer base and revenues.

And this is just the beginning.

To learn more about how RADCOM can help you leverage GenAI’s power, we invite you to attend our upcoming webinar, Talk the Talk: Why GenAI for Telcos. To register, click here.

This article was published on FutureNet, World Service Assurance & AI Insights blog.

Share this article
Skip to content