Enterprise Knowledge Search · IT Operations Automation
Telefónica's network operations teams had to handle growing network complexity, massive alarm volumes, and incident analysis workflows that made root-cause identification and troubleshooting slower and harder for NOC operators.
Reported impact
The initiative improves technician and engineer productivity, speeds incident analysis and troubleshooting, reduces operational-response times such as MTTD, MTTR, and MTNA, strengthens knowledge transfer and training for NOC engineers, and advances Telefónica's network-autonomy goals.
Telefónica Germany built the Network Operation Agent (NOA) as an internal generative AI troubleshooting assistant for network incident analysis. The application serves NOC operators through a chatbot interface, letting them ask natural-language questions about active incidents and receive troubleshooting guidance. On the ingest side, NOA pulls context from a database, operational documentation, live operational data, and the resolution history of previous incidents retrieved through an integrated ticketing system. In the processing tier, the source says the system uses LLM-based agents, reasoning mechanisms, RAG, and vectorization to correlate alarms, compare current incidents with prior cases, and identify likely root causes across domains. The resulting answers are delivered back to operators as precise natural-language responses, with the architecture intended to evolve further by adding orchestrators that can trigger intents and future closed-loop operational actions directly over the network.
NOC operators submit incident questions to NOA, which gathers context from a database, operational documentation, live operational data, and closed-issue records from the ticketing system. It applies RAG, vectorization, and reasoning to correlate alarms, identify related incidents and root causes, and returns natural-language troubleshooting guidance back to operators.
"In Telefonica Germany, they are developing several initiatives based on LLM and AI-agents for incident analysis that can help to the NOC operators to analyze and resolve the incidence."
"Is possible to review the information from a database, documentation and even from the resolution of previous incidents. It is integrated with the ticketing system and obtain information about the resolution of closed issues."
"NOA is not just a classic RAG-based chatbot solution. It is an application that combines reasoning mechanisms with direct connections to live operational data and knowledge documents enabling a much deeper understanding of network situations."
"Operating the network is becoming more complex every day due to the huge number of elements and the massive generation of alarms."
"Reduce times from the operational process like Mean Time to Detect (MTTD), Mean Time to Recover (MTTR) and Mean Time to Next Action (MTNA)."
"Within this topic, in Telefonica Germany have developed the Network Operation Agent (NOA)."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
In-house Sample Jul 23, 2026 Telefónica Telefónica needs to evolve and maintain a very large API and microservices estate without slowing delivery, while keeping integrations understandable, compliant, secure and maintainable; it also wants to reduce manual… AI agent · generative AI In-house Sample Sep 7, 2023 AT&T AT&T wanted to put generative AI into everyday employee workflows so software developers, engineers, finance teams, supply chain staff, HR, customer service, and fraud teams could work faster on code, data analysis,… in some cases 2x the speed, 1/2 the cost Ask AT&T · Falcon transformers · Meta LLaMA 2 
Microsoft Sample Nov 18, 2025 AT&T AT&T's employees were spending too much time investigating across many search links to find answers, developers were debugging code line by line, and data flowed faster than humans could keep up. Leadership needed proof… 100,000+ employees actively using agents Azure AI Foundry · Azure AI Search · Azure API Management 