Agriculture is a crucial aspect of our lives as it is a major source of employment and food. Soil is an important aspect of agriculture it is a source of structural support, provides nutrients to plants, regulates climate, etc. However, if soil composition is not maintained and controlled properly it could lead to low crop yield levels. The fertility of soil depends on various factors such as Organic matter content, pH, macronutrients (Magnesium Mg, Calcium Ca), and micronutrients (Zinc Zn, Iron Fe, Sodium Na, Copper Cu). Studying these factors helps determine the soil’s health. In the proposed paper we investigate the predictive power of machine learning algorithms in analyzing the soil fertility dependent on range of factors and its correlation with vegetation cover. By testing techniques such as Linear Regression, K-Nearest Neighbors (KNN), Decision Trees (DT), Ensemble Methods (Bagging), Random Forest (RF), and Gradient Boosting (GB) regressors on the dataset we predict soil fertility levels. Additionally, we integrate TPOT (Tree-based Pipeline Optimization Tool) to enhance the predictive power of these algorithms, ensuring improved accuracy and reliability in predictions.