Agriculture plays a significant role in human civilization, and crop yield depends on proper crop selection. Machine learning techniques can provide recommendation systems to guide farmers. This study applies the Light GBM algorithm to improve prediction accuracy for crop recommendations based on environmental factors and soil nutrient concentrations (N, P, K), as well as temperature, humidity, rainfall, and pH. The system also suggests fertilizers and predicts crop diseases to enhance agricultural productivity. © 2023 Elsevier B.V., All rights reserved.