关于构建基于检索增强生成的聊天机器人的FACTS研究
FACTS About Building Retrieval Augmented Generation-based Chatbots
AI总结：
本文针对企业级RAG聊天机器人构建的工程挑战，基于三款NVIDIA聊天机器人开发经验提出FACTS框架，给出15个RAG流水线控制点及大小LLM准确率-延迟权衡的实证结果，为构建安全企业级聊天机器人提供全面指导。
摘要：
由生成式人工智能驱动的企业聊天机器人正成为提升员工生产力的关键应用。检索增强生成（RAG）、大语言模型（LLM）以及Langchain、Llamaindex等编排框架是构建这类聊天机器人的核心要素。然而，打造高效的企业聊天机器人颇具挑战，需要精细的RAG流水线工程设计，具体包括嵌入模型与LLM的微调、从向量数据库中提取文档、查询改写、结果重排序、提示词设计、遵循文档访问控制规则、提供简洁回复、附上参考文献、保护个人信息以及构建编排智能体。我们基于在三款NVIDIA聊天机器人（分别用于IT/HR福利、财务财报及通用内容场景）的开发经验，提出了一个构建基于RAG的聊天机器人的框架。本文的贡献有三点：提出FACTS框架（即新鲜度、架构、成本、测试、安全），介绍15个RAG流水线控制点，以及提供大小LLM之间准确率-延迟权衡的实证结果。据我们所知，这是首篇针对构建安全企业级聊天机器人的各类因素及解决方案提供全面视角的同类论文。
英文摘要：
Enterprise chatbots, powered by generative AI, are emerging as key applications to enhance employee productivity. Retrieval Augmented Generation (RAG), Large Language Models (LLMs), and orchestration frameworks like Langchain and Llamaindex are crucial for building these chatbots. However, creating effective enterprise chatbots is challenging and requires meticulous RAG pipeline engineering. This includes fine-tuning embeddings and LLMs, extracting documents from vector databases, rephrasing queries, reranking results, designing prompts, honoring document access controls, providing concise responses, including references, safeguarding personal information, and building orchestration agents. We present a framework for building RAG-based chatbots based on our experience with three NVIDIA chatbots: for IT/HR benefits, financial earnings, and general content. Our contributions are three-fold: introducing the FACTS framework (Freshness, Architectures, Cost, Testing, Security), presenting fifteen RAG pipeline control points, and providing empirical results on accuracy-latency tradeoffs between large and small LLMs. To the best of our knowledge, this is the first paper of its kind that provides a holistic view of the factors as well as solutions for building secure enterprise-grade chatbots."
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
