AI-Powered Retail Insights and Query System

The dynamic retail industry demands efficient and intelligent solutions to address operational challenges and enhance customer experience. This paper presents an AI- powered system equipped with three core functionalities: (1) Retail shelf analysis using YOLOv5 to detect empty shelf spaces or identify products, (2) A text-to-query interface allowing users to upload database files and perform SQL queries using natural language, and (3) Sentiment analysis to extract valuable insights from customer feedback. The system is deployed via Streamlit and leverages CI/CD pipelines on GitHub and AWS EC2 for scalable, reliable performance. This study delves into the methodologies, implementation strategies, and the system’s transformative potential in modern retail operations. Keywords— Retail AI, YOLOv5, Text-to-Query System, Sentiment Analysis, Machine Learning, Automation