Krushimate-Smart Farming

Abstract

Agriculture remains a critical sector for economic stability and food security, yet crop diseases continue to cause significant yield losses worldwide. Early detection and timely advisory support are essential for mitigating these losses, particularly for small-scale farmers with limited access to agricultural experts. This paper presents a web-based smart farming application integrated with a large language model (LLM)–powered intelligent agent for crop disease analysis and decision support. The proposed system enables farmers to upload crop images for AI-assisted disease interpretation and interact with an intelligent agent to obtain practical guidance on disease severity, treatment, and prevention. The application is implemented using a lightweight Flask-based architecture and cloud-hosted AI inference, ensuring scalability and accessibility. Experimental validation demonstrates that the system provides real-time, farmer-friendly advisory responses, making it a cost-effective and deployable solution for precision agriculture.

KEYWORDS: Smart Farming, Crop Disease Detection, Intelligent Agent, Large Language Models, Precision Agriculture, Web Application