Using multi-agent architecture to mitigate the risk of LLM hallucinations

TL;DR

This paper proposes a multi-agent system combining LLMs and fuzzy logic to reduce hallucination risks in customer service applications via SMS.

Contribution

It introduces a novel multi-agent architecture that integrates LLMs with fuzzy logic specifically to mitigate hallucination risks in customer service.

Findings

Effective reduction of hallucination incidents in the system

Improved response accuracy in customer interactions

Demonstrated feasibility of multi-agent approach for risk mitigation

Abstract

Improving customer service quality and response time are critical factors for maintaining customer loyalty and increasing a company's market share. While adopting emerging technologies such as Large Language Models (LLMs) is becoming a necessity to achieve these goals, the risk of hallucination remains a major challenge. In this paper, we present a multi-agent system to handle customer requests sent via SMS. This system integrates LLM based agents with fuzzy logic to mitigate hallucination risks.