Modern farming must make decisions based on precise data in order to increase crop yields and reduce losses. The quality and amount of the harvest can be significantly impacted by an imbalance of nutrients in the soil, improper fertilizer use, and delayed detection of crop diseases. This study suggests an integrated machine learning framework with three main components: plant disease detection, fertilizer & crop recommendation, & soil nutrient prediction. In order to categorize soil fertility levels, the soil prediction model examines characteristics like pH, temperature, moisture, nitrogen, phosphorus, and potassium (NPK). Using a hybrid rule-based and supervised learning approach, the fertilizer recommendation module recommends the best crops & fertilizers based on anticipated soil properties. Plant leaf photos are categorized into healthy and diseased groups using deep learning in the disease detection model. For farmers, the combined system offers complete decision support. High prediction accuracy in all three modules is demonstrated by experimental results, indicating the efficacy of the integrated approach. The suggested framework can improve sustainable farming methods, decrease resource waste, and improve precision agriculture. Keywords—Soil Prediction, Fertilizer Recommendation, Crop Recommendation, Plant Disease Detection, Learning, Deep Learning, Precision Agriculture