Agrisense AI in Agriculture

Agricultural decisions depend on timely and locally relevant information about crop selection, soil health, irrigation, plant disease, weather, market prices, and government schemes. These services are often distributed across separate sources, creating access and coordination challenges for farmers. This paper presents AgriSense AI, a web-based farmer-support platform designed to bring these functions together through a single interface. The system combines farmer registration and profiles, crop recommendations, leaf-image disease assessment, soil-health analysis, irrigation guidance, weather and market information, scheme discovery, and an administrative dashboard. A bilingual conversational assistant, Krishi Mitra, is included to support English and Kannada interactions. The application follows a React and Vite front end with a Node.js and Express REST API, MongoDB for persistent records, and a locally hosted Ollama model for assistant responses. External weather data can be obtained through OpenWeatherMap when an API key is configured. The report describes a working prototype and its design, implementation, and test approach; it does not establish field-level agronomic accuracy or measured improvements in farmer outcomes. The current disease assessment and crop recommendation components have documented constraints, so their outputs should be treated as decision support rather than verified agronomic advice.