Deutsche Telekom is pushing agentic AI deeper into network operations.
The operator already runs RAN Guardian in Germany. It now plans broader automation through MINDR, a multi-agent network platform. Agentic AI means software agents can decide and act within approved limits.
The move targets a familiar telecom challenge. Mobile traffic can change quickly during concerts, sports events, or public incidents. Static planning often reacts too late.
Armin Sumesgutner, chief network officer at Deutsche Telekom, explained the issue clearly. “The most important issue is we are always behind the wave,” he said.
RAN Guardian addresses this with three cooperating agents. One scans public sources for events. Another studies the likely network impact. A third can apply selected fixes before congestion appears.
Those actions may include carrier aggregation and uplink adjustments. It can also change antenna tilt to focus capacity. These are practical tools for managing dense crowds.
This approach raises operational speed. Deutsche Telekom said the system can analyze 100,000 events. Previously, teams could handle around 1,000 events manually. The operator also wants to prevent 24 million poor-experience minutes.
However, automated network changes need careful control. A wrong adjustment can affect nearby users. It can also shift congestion instead of solving it. That makes guardrails and engineering oversight essential.
Sumesgutner said the system already handles many cases alone. “Most of the time, we fortunately already see congestions could be mitigated automatically, so full autonomous activity,” Sumesgutner said. “Some of them need some human interaction to make sure we’re going to solve it.”
Meanwhile, MINDR expands the concept beyond one network domain. It aims to connect signals from radio, transport, core, and service layers. That broader view matters because faults often cross several systems.
Arash Ashouriha, group technology officer, positioned MINDR as the next automation step. “RAN Guardian is a great example how far we already have progressed in our automation journey,” he said.
He added that current automation often remains too narrow. “Agentic AI will be a game changer because it allows us to move from individual automation use cases towards scalable autonomy,” Ashouriha said.
MINDR follows four main steps. It first correlates network signals. Then it uses root-cause analytics to find the issue. Root-cause analytics means identifying the real source of a fault.
After diagnosis, MINDR can execute and verify fixes. Less risky issues may run automatically. Critical changes still need human approval before execution.
That balance will define operator trust in autonomous networks. Speed brings clear value, especially during outages. Yet telecom networks carry critical services, businesses, and emergency communications.
Deutsche Telekom targets more than 90% automated ticket handling. It also wants issue qualification below 15 minutes. Incident resolution should fall under one hour.
The company plans an initial MINDR use case in Germany by late 2026. Wider deployment could follow across Germany and other markets. For engineers, the message is clear: automation is moving from scripts to decision-making systems.
