The exponential rise in structured data generation has created an unprecedented opportunity for data-driven decision-making. However, extracting actionable intelligence from raw CSV datasets remains a significant bottleneck for non-technical users. Traditional Exploratory Data Analysis (EDA) systems require specialized programming knowledge or complex Business Intelligence (BI) software. This paper proposes AURA Analyst (Autonomous Unified Reasoning Agent), a novel full-stack data analysis platform that simultaneously automates statistical profiling, natural language narrative generation, and time-series forecasting. We integrate a FastAPI-driven statistical engine (Pandas, NumPy) with advanced Large Language Models (Google Gemini, AWS Bedrock Claude 4.6) to provide zero-expertise insights. Experimental results demonstrate an end-to-end processing latency of under 10 seconds, 95% accuracy in automatic context classification, and robust 7-day forecasting capabilities using SARIMAX. The system is deployed as an interactive, terminal-themed web application featuring conversational data interrogation.
Patel et al. (Thu,) studied this question.