Deutsche Telekom has officially launched its AI-driven network management system, the RAN Guardian Agent, marking a major milestone in intelligent mobile network operations.



Deutsche Telekom jointly with Google Cloud at this year’s Mobile World Congress (MWC), the RAN Guardian Agent is designed to monitor network performance, support troubleshooting, and optimize operations in real time, ensuring faster responses and a more resilient network for customers.

AI-Driven Network Management
Abdu Mudesir, Board Member for Product and Technology at Deutsche Telekom, highlighted the innovation: “With the introduction of the RAN Guardian Agent, we are the first network operator to rely on a highly developed AI agent in network management. Our development teams have done pioneering work in the development of AI agents for mobile networks. By enabling intelligent interaction between our network experts and AI, we are solving specific challenges for the benefit of our customers, paving the way for autonomous, self-healing networks.”

The RAN Guardian Agent functions as a multi-agent AI system, with several specialized agents collaborating to ensure optimal network performance. It continuously analyzes network behavior, detects performance anomalies, and autonomously initiates corrective measures, enhancing both reliability and user experience.

How RAN Guardian Works
The agent operates in three key steps:

Event Analysis: The AI continuously scans publicly available sources, including social media and event directories, to identify upcoming public events in Germany. It assesses their scale, location, and potential network impact, verifying the data automatically.

Network Assessment: Another component evaluates the network’s capacity to handle increased traffic, monitors real-time parameters, and recognizes high utilization levels, recommending optimization measures when needed.

Autonomous Optimization: A separate AI agent executes these recommendations, reallocating resources or adjusting configurations. All actions are documented to continuously improve network readiness for future events.

Enhancing Network Efficiency
With mobile networks growing increasingly complex, traditional automation methods are no longer sufficient for real-time problem solving. Powered by large language models (LLMs) and advanced reasoning frameworks, the RAN Guardian Agent represents the next generation of intelligent automation, capable of understanding, reasoning, acting, and learning. Tasks that once took an hour can now be completed in mere minutes, significantly accelerating troubleshooting and maintaining consistently high mobile performance.

Technical Framework
The RAN Guardian Agent leverages Google Gemini 2.5, along with CloudRun, BigQuery, and Firestore, to deliver autonomous network management. Its key functionalities include:

Autonomous RAN Performance Monitoring: Continuous real-time analysis of network parameters to predict and detect anomalies.

AI-Driven Problem Classification and Routing: Prioritization of network degradations using multiple data sources, including monitoring, inventory, performance, and coverage data.

Proactive Network Optimization: Automated implementation of corrective actions, such as resource reallocation or network configuration adjustments.

By combining AI intelligence with human expertise, Deutsche Telekom’s RAN Guardian Agent is shaping the future of self-healing, high-performance mobile networks, ensuring faster response times, optimized operations, and superior customer experience across Germany.

Baburajan Kizhakedath