Supply Chain Resilience System using AI and ML

This paper presents a Supply Chain Resilience System that uses Artificial Intelligence (AI) and Machine Learning (ML) to help managers make better inventory and risk decisions. The system goes beyond simple sales forecasts by taking into account real-world external factors like geopolitical instability and severe weather. It uses an XGBoost-based prediction engine to classify the inventory state into three categories: Stock-Out Risk, Excess Inventory Risk, or Low Risk. Once a risk is detected, the system immediately gives a specific, numbered action, such as exactly how many units to order. The Gemini API is used to turn the numerical results into a clear, easy-to-read executive summary. The platform also includes an Explainable AI (XAI) feature that shows which input factors contributed most to the risk score, making the system transparent and trustworthy. A role-based dashboard separates the experience for executives and analysts, so each user gets the right information for their job. All prediction records are saved to Firebase Firestore, creating a useful history for future review. The system was built using Next.js for the frontend, FastAPI for the backend, and MongoDB Atlas for data storage. Testing showed that the system delivers predictions in under one second and handles all major workflows reliably.