Magyar Telekom is shifting away from a traditional, subscriber-report-driven fault response model toward a more predictive, increasingly AI-assisted network monitoring approach. Csutorka Gábor, head of service and network operations at Magyar Telekom, notes that when a customer calls in, it is often already too late because something in the system has gone wrong.
"We must be fundamentally capable of extracting information directly from the network," Csutorka said. The company aims to use network-derived data and automated analysis to detect early signs of issues and intervene before service outages or performance degradations become noticeable.
Why the change?
The rationale is that maintaining service quality is more effective when done predictively rather than reactively. Predictive operations can reduce customer complaints related to network faults and shorten recovery times. Magyar Telekom says it is therefore integrating more data-driven and automated tools into operations, which continuously process network telemetry, alerts, and performance metrics.
Role of automation and artificial intelligence
Automation and artificial intelligence (AI) play a significant role in the plans: algorithms help identify patterns, forecast emerging faults, and either recommend or automatically execute corrective actions. According to Csutorka Gábor, the ideal state is one where the network itself provides the information needed for operations to act before a subscriber notices a problem.
What this means for subscribers
The benefit for subscribers is improved reliability: fewer outages, faster fixes, and less manual intervention. The transition is gradual: Magyar Telekom is progressively adding more data sources and automated solutions to operations while developing predictive practices based on network data analysis.
Summary
Magyar Telekom's network operations are moving from reactive fault handling toward predictive, data-driven, and partly AI-based oversight. The shift is intended to allow the operator to prevent disruptions by relying on information extracted from the network rather than discovering problems through subscriber reports.
