Home > Insights > Autonomous Network Operations: The Journey to L4/5

Untangling the Hype: How to Maximize the Power of Agentic AI in Telecom Networks

Agentic AI in Telecoms, Autonomous Network Operations: The Journey to L4/5

An article by Michal Fridman, VP Marketing and Business Development, Radcom.

Agentic AI promises to reshape telecom networks. With the advancement in AI at an unprecedented pace, telecom operators can expect dramatically improved processes and efficiency, potentially changing the fundamental way they work. Where traditional AI sticks to pre-programmed rules or optimization within a set framework and generative AI (GenAI) emits outputs and directions, agentic AI allows the models to understand deeper and adapt dynamically.

How it works

Agentic AI involves combining multiple AI models that understand and interpret the data and react to predefined tasks. It doesn’t just suggest an outcome, but rather decides the best way to achieve the desired outcome. In the telecom context, it has the potential to transform the network into an autonomous system – managing complex tasks, understanding priorities and proactively making adjustments. And as the technology progresses, these advanced capabilities will allow AI agents to collect data from numerous sources and independently act in real-time with minimal human supervision.

It’s not all hype: benefits of agentic AI

It is no wonder that agentic AI is expected to take the world by storm, growing to USD 251.8 billion by 2034 from USD 3.8 billion in 2024, as it is anticipated to play a significant role in efficiency and cost savings. One study quantified the benefits of agentic AI, between $10-20 million in an organization of 1,000 employees, based on offloading 12.5% of human work. Another study showed that implementing advanced, responsible AI frameworks could translate into $250 billion in value for telcos worldwide by 2040.

Customer experience is another area where agentic AI is projected to alleviate current quandaries. A recent survey shows that almost half of telecom professionals are investing in AI mainly for customer experience optimization. By understanding issues and priorities, AI agents are expected to enhance network performance, automate troubleshooting, and improve customer experience. Communications services providers (CSPs) are likely to see a transformation in key areas such as: i) customer care, ii) customer experience and iii) network operations.

Customer care will be conducted through smarter virtual assistants that can proactively react to customer issues, fulfill individual requests or provide complex information to the subscriber. Customer experience will be improved as technologies can predict preferences to autonomously carry out tasks to boost service and enhance subscriber experience. Network operations will be optimized with automation that understands the root cause of a problem, adjusts the network and predicts quality lapses.

The challenge of reliable data

Like generative AI, however, its success will depend largely on the reliability of the data. If AI agents are fed “garbage in,” then they will produce “garbage out.” The accuracy of the output relies on valuable input. This means the ability to make decisions demands accurate, complete, and reliable data.

Telecom operators sit on large amounts of data, making them a primary candidate to benefit from agentic AI. On the other hand, much of the data is stuck in silos, and without a clear strategy to bring it all together, there’s no reliable single source of truth.

Accessing all the data is the first hurdle. In telecom operations, AI agents need precise data about everything from network configurations to subscriber-level analytics to formulate reliable and trustworthy decisions or solutions. Once the problem of accessing data is solved, only a limited amount of data on network errors exists, which can risk imbalanced datasets. This can result in bias, incomplete data, or ineffective decisions from the AI agent.

Look after your network, to look after your customers

The impact on network operations and customer interactions in these cases can be enormous, resulting in lasting consequences such as loss of subscribers or more serious degradations and anomalies on the network. This will have the opposite effect on network optimization and quality customer experience.

Moving into an agentic AI and more futuristic artificial general intelligence era, ensuring data accuracy and validity will be more critical than ever. This is where the next generation of automated assurance solutions come in. Assurance solutions already analyze all the data on the network from end to end, bringing together insights from multiple, siloed sources. Combining that with agentic AI capabilities will bring the next level of intelligent assurance that is the key to ensuring customer experiences.

Using telco-trained traditional and generative AI models, intelligent assurance solutions utilize the existing data on the operator’s network. This helps balance datasets, ensuring AI decisions are accurate and anonymized to produce reliable and trustworthy insights and offers access to real-time balanced data models.

Intelligent assurance solutions are designed to sustain millions of connected devices, high-speed voice and data services, and changing conditions to provide insights into user analytics, location intelligence, usage patterns, service interactions, and more. This allows operators to identify trends, locate and predict potential faults, and offer premium services with superior customer experience.

These solutions also shine a spotlight on the customer. They process full, granular-level subscriber data and offer automated root-cause analysis. They also provide operators with specialized agents to ensure customer engagement, superior service performance and even service design and orchestration. The adeptness of agentic AI to collect data from different sources also means that assurance solutions can be integrated with other management systems to prioritize the subscriber experience.

In short: beware of the data

At the pace AI is moving, we can expect the unimaginable to become the norm—but only with the right type of data. Combining agentic AI with next-generation assurance offers reliable data and predictive insights. This, in turn, transforms the network into a proactive, preemptive system, providing customers with hyper-personalized services and the ultimate premium experience—whether it is from generative AI, agentic AI, or next-generation artificial general intelligence.