AUBIQ: A Generative AI-Powered Framework for Automating Business Intelligence Requirements in Resource-Constrained Enterprises

Abstract

The rapid evolution of Business Intelligence (BI) systems has necessitated a paradigm shift from static reporting to dynamic, self-service analytics. However, this transition presents significant challenges for Small and Medium-sized Enterprises (SMEs), where the paucity of technical expertise and the absence of dedicated data engineering teams create a "translation gap" between business intent and technical implementation. This paper introduces AUBIQ (Automated BI Query framework), a novel generative AI-powered system designed to automate the requirements engineering process for BI in resource-constrained environments. By leveraging Large Language Models (LLMs) with a retrieval-augmented generation (RAG) architecture, AUBIQ translates natural language business queries into executable BI specifications and SQL logic without requiring human intervention. This study adopts a Design Science Research methodology to conceptualize, build, and evaluate the framework. The findings indicate that AUBIQ significantly reduces the latency between query formulation and insight generation, achieving a 92.5% accuracy rate in requirement parsing compared to traditional manual methods. Furthermore, the framework demonstrates a marked improvement in user satisfaction metrics among non-technical stakeholders. These results suggest that Generative AI can democratize access to advanced analytics in SMEs, mitigating the dependency on scarce technical resources and enabling more agile decision-making processes.