KrushiLINK: AN AI-POWERED SMART NURSERY SELECTION, SEEDLING BOOKING, AND CROP HEALTH ADVISORY PLATFORM

Small and marginal farmers typically rely on scattered information sources while choosing the planting material, locating a suitable nursery, diagnosing crop diseases, and purchasing agricultural input. Previous works in digital agriculture have demonstrated crop recommendation, soil analysis, disease recognition, agricultural marketplaces, and farmer-assistant systems, but these features are usually offered as stand-alone applications or limited research prototypes. This paper presents KrushiLINK, an AI-powered platform connecting farmers with nurseries and agricultural input suppliers through an end-to-end system. The platform enables farmers and nursery owners to browse, book, manage, and locate seedlings and agricultural input while implementing a machine-learning recommendation system for recommending nurseries based on crop, location, land size, quantity, budget, season, and available agricultural parameters. The paper introduces a novel crop-health module allowing a farmer to upload an image of a crop leaf, process the input with a CNN or transfer-learning-based classifier to identify crop diseases, and cross-reference the problem with a treatment knowledge repository. The platform then recommends local fertilizer/agri-input shops and their products in case inventory information is available on the marketplace. The paper describes the system design, recommendation model, crop disease detection pipeline, data model, research gap, and an evaluation methodology relying on accuracy, precision, recall, F1, recommendation relevance, response time, and usability. The work focuses on a decision-support system, which should not be mistaken for a diagnostic tool, as treatment recommendations should be evaluated by cross-referencing with crop-specific guidelines, formulation instructions, regulatory rules, and agricultural expert advice.