跳转至

文章背景与核心概要

传统的检索增强生成（RAG）系统在客户联络中心等实时对话场景中，常面临查询构建不准确以及对常见问题（FAQ）进行冗余检索等挑战。本文介绍了一种名为“Beyond-RAG”的决策支持系统，旨在实时识别客户提出的问题。如果查询与现有FAQ匹配，系统将直接检索答案；否则，将启用RAG机制生成回答。

该系统已在 Minerva CQ 成功部署，能够在 2 秒内向人工客服提供响应，从而显著缩短平均处理时间（AHT）并降低运营成本。此外，针对缺乏预定义FAQ的历史对话记录，作者还引入了一种自动化的LLM智能体工作流来从中提取FAQ，为智能客服系统的工程落地提供了极具价值的参考方案。

Beyond-RAG: Question Identification and Answer Generation in Real-Time Conversations

Beyond-RAG: Question Identification and Answer Generation in Real-Time Conversations

📌 Summary

📌 Summary

Traditional Retrieval-Augmented Generation (RAG) systems used in customer contact centers often struggle with real-time conversations due to challenges like inaccurate query formulation and redundant retrieval of Frequently Asked Questions (FAQs). This paper introduces Beyond-RAG, a decision support system designed to identify customer questions in real time. If a query matches an FAQ, the system retrieves the answer directly; otherwise, it utilizes RAG. Deployed at Minerva CQ, the solution delivers responses to human agents within 2 seconds, reducing Average Handling Time (AHT) and operational costs. Additionally, the authors introduce an automated LLM-agentic workflow to extract FAQs from historical transcripts when predefined FAQs are absent.

📄 Article Metadata

📄 Article Metadata

  • arXiv ID: arXiv:2410.10136 [cs.CL]
  • Related DOI: 10.24251/HICSS.2026.213
  • Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
  • Authors: Garima Agrawal, Sashank Gummuluri, Cosimo Spera
  • Submitted On: 14 Oct 2024
  • Last Revised: 1 Sep 2026 (Version v2)

📝 Abstract

在客户联络中心中，由于人工客服需要手动解读客户查询并检索相关的知识库（KB）文章，往往会导致较长的平均处理时间（AHT）。尽管利用大语言模型（LLMs）的检索增强生成（RAG）系统已被业界广泛采用以协助处理此类任务，但RAG在实时对话中仍面临诸多挑战，例如查询构建不准确以及对常见问题（FAQ）的冗余检索。

In customer contact centers, human agents often struggle with long average handling times (AHT) due to the need to manually interpret queries and retrieve relevant knowledge base (KB) articles. While retrieval augmented generation (RAG) systems using large language models (LLMs) have been widely adopted in industry to assist with such tasks, RAG faces challenges in real-time conversations, such as inaccurate query formulation and redundant retrieval of frequently asked questions (FAQs).

为了解决这些局限性，我们提出了一种决策支持系统，该系统通过首先实时识别客户问题来超越传统的RAG。如果查询与FAQ匹配，系统直接从FAQ数据库中检索答案；否则，通过RAG生成答案。我们的方法减少了对手动查询的依赖，能够在 2 秒内向客服提供响应。该系统部署在Minerva CQ的人工客服AI辅助解决方案中，能够提高效率、缩短AHT并降低运营成本。此外，当不存在预定义FAQ时，我们还引入了一种自动化的LLM智能体工作流，用于从历史对话记录中识别FAQ。

To address these limitations, we propose a decision support system that can look beyond RAG by first identifying customer questions in real time. If the query matches an FAQ, the system retrieves the answer directly from the FAQ database; otherwise, it generates answers via RAG. Our approach reduces reliance on manual queries, providing responses to agents within 2 seconds. Deployed in AI-powered human-agent assist solution at Minerva CQ, this system improves efficiency, reduces AHT, and lowers operational costs. We also introduce an automated LLM-agentic workflow to identify FAQs from historical transcripts when no predefined FAQs exist.