Summary
Troubleshooting optical networks can be tricky because AI systems sometimes make up data, leading to wrong conclusions. The authors show how adding real data sources into an AI agent helps stop this by making the AI’s reasoning more reliable. This leads to better identification of network problems and clearer explanations of what went wrong. The system aims to build trust by avoiding mistakes caused by AI hallucinations.
What this means in practice
- •For optical network operators: Use AI that references real data to accurately diagnose and explain network faults, reducing misdiagnosis from AI errors.
- •For it infrastructure teams: Improve automated root-cause analysis systems by integrating verified data to avoid false alarms caused by AI-generated incorrect information.
Authors
Jeremias Dötterl, Martin Kuipers
Abstract
An agentic AI system for automated troubleshooting in optical communication networks is demonstrated that integrates data references into the agent architecture to prevent data hallucinations, improving the interpretability and trustworthiness of root-cause identification and reasoning.