• Jasmitha D
• Jeevitha N N
• Meghana R
• Monisha Shrikant Naik

The mission of this endeavor is to design an integrated AI-based platform for farmers that enhances farmers with wise suggestions and transparent connectivity with customers and administrators. Conventional agriculture systems usually are deprived of the timely availability of market data, efficient crop handling tools, and efficient distribution routes. To meet these requirements, this project proposes FarmLink AI, a system architecture that integrates machine learning methodologies like Decision Tree Classifiers, Random Forest ensembles, and Convolutional Neural Networks (CNNs) to enhance agricultural productivity and promote sustainable farming practices. The platform consists of three central modules: Admin, Farmer, and Customer. It comprises components like crop selection according to climate conditions, fertilizer recommendations according to soil condition, and disease prediction using image-based CNN models. The model was trained and tested on actual-world datasets concerning crop types, soil parameters, and images of diseased leaves. The outcome reveals that the AI models provide high accuracy in decision support tasks. This work depicts the ability of machine learning and deep learning to transform agriculture into a technologically advanced discipline that makes smarter decisions in farming and efficient farm-to-market communication.

Machine Learning, CNN, Decision Tree, Random Forest, Flask, MongoDB, Fertilizer Recommendation, Crop Disease Detection, Crop Recommendation, Smart Farming, Precision Agriculture

"FARMLINK AI", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 7, page no.472-476, July-2025, Available :http://www.jetir.org/papers/JETIRGX06089.pdf

"FARMLINK AI", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 7, page no. pp472-476, July-2025, Available at : http://www.jetir.org/papers/JETIRGX06089.pdf