Authors: Kale Shruti Sunil, Professor Bangar A.P, Professor Bhosale S.B., Professor Karad V. A
Abstract: Food and medicines that pass their expiry dates cause avoidable financial loss, health risk and environmental harm, yet most expiry-tracking tools are passive reminder lists that depend on manual data entry. This paper proposes an Agentic AIPowered Expiry Date Management System in which a team of cooperating, tool-using AI agents autonomously captures product labels, extracts and validates expiry dates with optical character recognition (OCR), estimates shelf life when no date is printed, ranks items by spoilage risk, and acts on the result by sending timely alerts, suggesting recipes and routing surplus items to donation partners. A central orchestrator agent plans tasks, calls specialised agents and learns from user feedback, while a shared knowledge layer keeps inventory and food-safety rules consistent. We describe the architecture, the riskscoring model and the agent workflow, and report an illustrative simulation of 500 households over 60 days in which the agentic prioritisation policy reduced wasted items from 21.5% to 14.8% (c = 0.7) and 13.3% (c = 0.9). The modular design allows each agent to be upgraded or deployed independently in households, retail shelves and hospital pharmacies.
