In order to ensure the protection of food, sustainable agriculture techniques are essential given the expanding global population and growing demand for food. Efficient crop selection and effective use of fertilizers are essential for maximizing productivity and reducing environmental impact. GrowWise is a web-based tool offered by this observation that uses machine learning techniques to help farmers determine the best crop to plant and the optimal amount of fertilizer to use based on the nutritional composition of the soil. To improve model performance, our approach includes thorough data exploration, visualization, preprocessing, and feature engineering. The use of K-fold cross-validation to provide robustness and eliminate outliers is done, in order to address problems with data quality. Several performance metrics were used to assess the created models: mean squared error was used for soil nutrient composition and accuracy, F1-score, precision, recall, and support was used for crop and fertilizer recommendation. The outcomes show GrowWise's ability to support sustainable farming practices and higher crop yields by offering precise advice on crop selection and fertilizer application. This study shows how machine learning (ML) may be used to create user-friendly decision support systems like GrowWise. These systems have the potential to greatly improve farmers’ ability to make decisions, optimize the use of their resources, and ultimately raise agricultural productivity.