Retrieval-Augmented Generation (RAG) improves the accuracy and relevance of large language model outputs by incorporating knowledge retrieval. However, implementing RAG in enterprises poses challenges around data security, accuracy, scalability, and integration. This paper explores the unique requirements for enterprise RAG, surveys current approaches and limitations, and discusses potential advances in semantic search, hybrid queries, and optimized retrieval. It proposes an evaluation framework to validate enterprise RAG solutions, including quantitative testing, qualitative analysis, ablation studies, and industry case studies. This framework aims to help demonstrate the ability of purpose-built RAG architectures to deliver accuracy and relevance improvements with enterprise-grade security, compliance and integration. The paper concludes with implications for enterprise deployments, limitations, and future research directions. Close collaboration between researchers and industry partners may accelerate progress in developing and deploying retrieval-augmented generation technology.

通过结合知识检索，改进大型语言模型的输出的准确性和相关性的检索增强生成（RAG）在企业中的实施面临数据安全、准确性、可扩展性和集成方面的挑战。本文探讨了企业 RAG 的独特需求，调查了当前方法和限制，并讨论了语义搜索、混合查询和优化检索的潜在进展。它提出了一个评估框架，以验证企业 RAG 解决方案的能力，包括定量测试、定性分析、消融研究和行业案例研究。该框架旨在帮助证明面向企业级安全、合规性和集成性能够提供准确性和相关性改进的目的构建的 RAG 架构的能力。本文总结了企业部署的含义、限制和未来研究方向。研究人员与行业合作伙伴之间的密切合作可能加速检索增强生成技术的开发和部署的进展。

RAG 不适用于企业