Cross-Domain Query Translation for Network Troubleshooting: A Multi-Agent LLM Framework with Privacy Preservation and Self-Reflection

TL;DR

This paper introduces a hierarchical multi-agent LLM framework for cross-domain query translation in network troubleshooting, focusing on privacy preservation, accurate classification, and user-friendly communication.

Contribution

It proposes a novel multi-agent LLM architecture with self-reflection, privacy-preserving anonymization, and few-shot learning for effective network troubleshooting across domains.

Findings

Achieved accurate query classification with a dual-stage hierarchical approach.

Maintained user privacy through semantic anonymization techniques.

Successfully translated technical responses into understandable language.

Abstract

This paper presents a hierarchical multi-agent LLM architecture to bridge communication gaps between non-technical end users and telecommunications domain experts in private network environments. We propose a cross-domain query translation framework that leverages specialized language models coordinated through multi-agent reflection-based reasoning. The resulting system addresses three critical challenges: (1) accurately classify user queries related to telecommunications network issues using a dual-stage hierarchical approach, (2) preserve user privacy through the anonymization of semantically relevant personally identifiable information (PII) while maintaining diagnostic utility, and (3) translate technical expert responses into user-comprehensible language. Our approach employs ReAct-style agents enhanced with self-reflection mechanisms for iterative output refinement, semantic-preserving anonymization techniques respecting $k$-anonymity and differential privacy principles, and few-shot learning strategies designed for limited training data scenarios. The framework was comprehensively evaluated on 10,000 previously unseen validation scenarios across various vertical industries.