Cloistered Knowledge Capture and Retrieval: Offline LLMs and Vector Search for Enterprise
Stephen J. Hall, Serengul Smith, Can Başķent, Clifford De Raffaele · 2025
In an era of rapid technological advancements, Artificial Intelligence (AI) has proven to be an invaluable tool for enhancing various aspects of human life. This research paper delves into the potential impact of cutting-edge AI technologies on business by integrating them into a knowledge management system. The aim of this study is to explore how these innovations can contribute significantly to empowering businesses, facilitating their growth, bolstering their competitiveness in the marketplace, and provide resilience against staff mobility. This paper examines recent advancements in Generative Transformers, Vector Databases, and Large Language Models to support the development of a novel, secure system for the capture, storage, and retrieval of knowledge—purpose-built for business-focused knowledge management. It implements and evaluates a system that challenges a sanitized version of an actual business dataset consisting of 595 documents to a set of 40 actual business enquiries. The accuracy, trustworthiness and response time of the system is evaluated with a thorough review of its limitations and recommendations for future work.