Farm Chatbot for Farmers Using Machine Learning and Optimization

Abstract

Theagriculturalsectorisanimportantpartofthesocio-economicdevelopmentofdevelopingnations.But farmers are unable to get personalized and real-time advisory services. This paper presents a Smart Agriculture Advisory System that combines a conversational AI chatbot with a supervised Machine Learning model implemented on a serverless cloud platform. The system is implemented using a React- basedfrontendandSupabaseBackend-as-a-Service(BaaS)forsecuredatamanagementwithRow-Level Security (RLS). A Random Forest Classifier is trained on soil nutrient and climatic factors to provide dynamic suggestions for crops. The trained model is implemented on a FastAPI-based microservice and incorporated into the chatbot flow. Experimental outcomes show high prediction accuracy and efficient user interaction performance compared to existing advisory systems. The proposed system offers a scalable, secure, and intelligent decision-support service for farmers.

Introduction

Agriculture is vital to India’s economy, but many farmers still rely on traditional methods, leading to low productivity and poor decision-making. Existing digital solutions lack personalization, security, and user-friendly design. To address these gaps, the proposed system integrates AI, cloud computing, and serverless architecture to deliver a scalable, secure, and intelligent advisory platform.

The system features an AI-based chatbot that provides real-time guidance on crop selection, weather updates, market prices, and government schemes, while using personalized farmer profiles for tailored recommendations. It is built as a full-stack web application with a React frontend and a Supabase backend, ensuring scalability, efficient data management, and secure multi-user access through JWT authentication and Row-Level Security (RLS).

A Random Forest machine learning model, trained on agricultural data (soil nutrients, weather, pH, etc.), predicts suitable crops with high accuracy (~96.3%) and is deployed via a FastAPI microservice. The architecture is modular, combining frontend, backend, database, and ML layers, enabling real-time predictions and efficient system performance.

Conclusion

This paper proposes an intelligent Smart Agriculture Advisory System that combines a trained Random Forest model for crop recommendation with a secure serverless cloud architecture. The system combines conversational AI, cloud data management, and supervised machine learning to offer personalized agriculturaladvice.ThecombinationofFastAPImodel deploymentandReactfrontendimplementationproves the viability of a scalable AI-driven agricultural advisorysystem.Theproposedarchitectureprovidesa solid foundation for future improvements such as IoT integration, predictive analytics, and sophisticated AI- driven decision support systems.

References

[1] FAO, “Digital Technologies in Agriculture,” Food and Agriculture Organization, 2022. [2] R. Rajesh et al., “Machine Learning in Crop Recommendation Systems,” IEEE Access, 2021. [3] Supabase Documentation, “Row-Level Security Policies,” 2024. [4] React Documentation, “React Official Guide,” Meta, 2024. [5] PostgreSQL Global Development Group, “PostgreSQL Documentation,” 2024. [6] M. Fowler, “Serverless Architectures,” IEEE Software, 2019. [7] S. Wolfert, L. Ge, C. Verdouw, and M. Bogaardt, “BigDatainSmartFarming–AReview,”Agricultural Systems, vol. 153, pp. 69–80, 2017. [8] R. Kamilaris and F. X. Prenafeta-Boldú, “Deep Learning in Agriculture: A Survey,” Computers and Electronics in Agriculture, vol. 147, pp. 70–90, 2018. [9] A. Khaki and L. Wang, “Crop Yield Prediction Using Deep Neural Networks,” Frontiers in Plant Science, vol. 10, 2019. [10] P. Sharma, A. Jain, and S. Gupta, “Machine Learning-Based Crop Recommendation System Using Soil and Climate Data,” IEEE Access, vol. 8, pp. 122345–122356, 2020. [11] G. Matthews, “Decision Support Systems in Agriculture,” Journal of Agricultural Informatics, vol. 9, no. 2, pp. 1–10, 2018. [12] J. Jones et al., “Toward a New Generation of Agricultural System Data, Models, and Knowledge Products,”AgriculturalSystems,vol.155,pp.269–288, 2017.

Copyright

Copyright © 2026 Dr. Deevi HarriKrishna, Sk. GhanSaida, Sk. Yasin, Sk. Nagoor Jani Basha, Sk. Mohammad Aman Ali. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.