Development and Implementation of a Retrieval-Augmented Generation System: The Lookinglass

This thesis proposes the development of Lookinglass, a Retrieval-Augmented Generation (RAG) system that combines language models and dynamic data retrieval for accurate and context-relevant responses. It aims to enhance information accuracy, enable versatile data handling, and improve decision-making efficiency in businesses. The RAG system integrates techniques for data processing, query understanding, response generation, and rigorous testing to transform data interaction experiences.