Deutsche Telekom has previously described MINDR as an evolution of RAN Guardian, extending agentic AI from domain-level optimization toward service-level intelligence across complex, multi-domain network environments
In sum – what to know
RAN Guardian scales – The live system uses three agents to identify events, assess network impact and autonomously implement selected remediation actions.
MINDR broadens scope – The multi-agent platform correlates network signals, diagnoses incidents, and executes remediation within predefined guardrails.
Level 4 target – Deutsche Telekom wants to move from individual automation use cases toward increasingly autonomous networks, while retaining human approval for critical changes.
Deutsche Telekom is expanding its use of agentic AI from radio access network (RAN) operations to broader network automation, with RAN Guardian already live in Germany and MINDR being developed as a cross-domain platform.
At Deutsche Telekom’s recent AI Investor Day, Armin Sumesgutner, chief network officer at Deutsche Telekom, presented RAN Guardian as an agent-based approach to preparing the mobile network for large events, while Arash Ashouriha, group technology officer, outlined MINDR as a broader platform for autonomous network operations.
RAN Guardian targets network demand around events
Sumesgutner said traditional network planning is largely static and can struggle when large numbers of people gather unexpectedly or when events disrupt normal traffic patterns. “The most important issue is we are always behind the wave,” he said.
RAN Guardian addresses this through three agents. The first has a scanning function, collecting information from public media, social media, and other sources to identify events. A monitoring agent then categorizes the information, assesses the technical implications, and evaluates the potential impact on the network The third component is the remediation agent, which can take action ahead of an event.
For events that could be handled through the system, the remediation agent could make changes including carrier aggregation, uplink and antenna-tilt adjustments to focus capacity where it was needed.
“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.”
Sumesgutner said the approach expands the number of events that can be analyzed and addressed. He compared the roughly 1,000 events that could previously be handled manually with 100,000 events that the agentic approach can analyze and remediate. He also said the system is aimed at avoiding 24 million minutes of poor customer experience.
Deutsche Telekom had already described RAN Guardian as a live multi-agent system capable of monitoring network behavior around events, assessing network capacity and initiating corrective actions. The company said the system went live in Germany in November 2025 and had autonomously triggered more than 100 remediation actions during its first month.
MINDR moves beyond individual network domains
Ashouriha positioned MINDR as the next step in Deutsche Telekom’s automation journey. “RAN Guardian is a great example how far we already have progressed in our automation journey,” he said. However, he said most existing automation remains domain-specific and focused on individual use cases. Deutsche Telekom’s ambition is to move toward Level 4 autonomous networks.
“Agentic AI will be a game changer because it allows us to move from individual automation use cases towards scalable autonomy,” Ashouriha said.
MINDR is designed as a multi-agent platform that brings specialized agents across the network together. “MINDR is a modular autonomous observability and incident management platform that brings specialized agents across the network together,” Ashouriha said. “They can anticipate issues, reason different systems, and increasingly take autonomous action.”
The platform operates through four main steps. First, it correlates signals from across the network. Root-cause analytics then uses specialized agents to diagnose issues across network domains and nodes.
The third step is remediation. Ashouriha said MINDR can execute and verify remediation within predefined guardrails. Less critical issues can be handled autonomously, while critical changes require human approval. “Some of the issues can be fixed and the impact practically is low. So, we may choose that the agent does it autonomously, but for every critical change, there will be always a human in the loop, which has to approve the change before it gets executed,” he said.
The fourth step is continuous learning from the outcome, allowing the platform to use previous incidents and their results to improve how future problems are handled.
Deutsche Telekom is targeting more than 90% automated handling of tickets, average issue qualification in less than 15 minutes and incident resolution below one hour. Ashouriha also said the company is targeting a 30% productivity improvement. “Autonomy cannot be just about efficiency. Ultimately it has to translate into a better customer experience,” he said.
Deutsche Telekom has previously described MINDR as an evolution of RAN Guardian, extending agentic AI from domain-level optimization toward service-level intelligence across complex, multi-domain network environments.
At the AI Investor Day, Ashouriha said Deutsche Telekom plans to introduce an initial MINDR use case as an MVP in Germany by the end of 2026, followed by broader scaling of the platform across Germany and other markets.
Table of Contents
