This paper presents SmartCrop, a comprehensive full-stack Precision Agriculture Decision Support System that integrates machine learning, generative artificial intelligence, and real-time environmental data to assist smallholder farmers in making informed agronomical decisions. The system is architected as three independent microservices: a React.js (Vite) frontend, a Node.js/Express.js REST API backend with MongoDB persistence, and a Python Flask-based machine learning inference service. The machine learning service employs a Random Forest Classifier trained on 2,200 soil and climate samples for crop recommendation across 22 crop varieties achieving 99.3% test accuracy, a Random Forest Regressor for yield forecasting with R² score of 0.983 and MAE of 0.24 tons/hectare, and a Random Forest Classifier for fertilizer recommendation achieving 93.4% test accuracy. Soil health analysis is enhanced through integration of Google Gemini 2.5 Flash generative AI, providing context-aware, dosage-specific fertilizer recommendations and crop disease risk assessments. The system additionally incorporates real-time weather intelligence via OpenWeatherMap API, smart irrigation scheduling, live market price analysis with financial profitability estimation, and an automated notification engine spanning five agricultural alert domains. Secure user authentication is implemented using JWT tokens with bcrypt password hashing, email verification, and password reset workflows. Experimental evaluation demonstrates 100% functional test case pass rate across 45 test cases, confirming practical viability for real-world agricultural deployment. Technologies: React.js, Node.js, Express.js, MongoDB, Python, Flask, scikit-learn, Google Gemini AI, OpenWeatherMap API, JWT, bcrypt, Chart.js, Mongoose, nodemailer.
Karan Ankade (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: