AI-Powered Smart Inventory Management: Enhancing Efficiency Through Predictive Analytics and Automation

Traditional inventory management systems suffer from inefficiencies, manual intervention, and high error rates. This paper presents an AI-driven smart inventory system using QR code scanning, machine learning (Random Forest, Linear Regression), and predictive analytics for real-time stock tracking and sales forecasting. A Raspberry Pi 5 with webcams enables automated product scanning, achieving 99.2% accuracy with 0.9s latency. Out of the two models Random Forest model indicated higher R-squared of 0.89 compared to the Linear regression model with R-squared of 0.68. The system improves inventory efficiency by predicting demand trends and alerting users about low stock levels. Security measures include QR code authentication, role-based access control, and AES-256 encryption to protect inventory data. Compared to traditional systems and commercial solutions like SAP and Oracle, our approach is cost-effective, scalable, and suitable for small to medium-sized businesses. Future work includes API integration with e-commerce platforms and warehouse automation.