Farmonic: A Decision-Driven AI-Integrated Agricultural E-Commerce Ecosystem

Jagappa , Ashish Khadela Maheshbhai, Srihari J, M Jessica, Mitinam Paloh, Ramesha K

Digital transformation in agriculture has largely progressed through fragmented solutions, including independent e-commerce platforms, advisory applications, and equipment rental services. Despite these advancements, existing systems fail to provide an integrated, intelligence-driven framework that connects crop diagnosis, contextual decision support, and transactional workflows within a unified ecosystem. This study proposes Farmonic, a decision-driven agricultural e-commerce architecture that seamlessly integrates AI-based plant disease detection, weather-aware crop advisory, and machinery rental management through a scalable microservice infrastructure. The proposed system employs transfer learning–based deep convolutional neural networks for multi-class crop disease classification and incorporates a context-aware recommendation engine that considers crop stage, environmental conditions, and diagnostic outputs to optimize product suggestions. A closed-loop AI-commerce pipeline is established, linking diagnosis, advisory, procurement, and post-transaction feedback into a continuous intelligent workflow. Experimental validation demonstrates a classification accuracy of 92%, with strong precision–recall performance and low system latency under simulated load conditions. The results confirm the feasibility of integrating artificial intelligence with agricultural commerce to enhance transparency, operational efficiency, and data-driven decision making. Farmonic advances the concept of outcome-centric digital agriculture by transforming traditional marketplaces into intelligent farming ecosystems that support the complete agricultural lifecycle.