Agriculture remains the primary occupation for a majority of the Indian population, yet granting much emphasis to subjective decision-making of traditional farming texts will lead to inefficiency, wastage of resources, and a decrease in crop yields. To mitigate these problems, there is an acute need for technology-based and data-oriented methods that may optimize agricultural practices for sustainable development. The growing demand for sustainable agricultural practices in the face of climate change, soil degradation, and rising food demand presents a significant challenge in India. Small and marginal farmers are almost never given timely and accurate advice on crops and fertilizers, for which the farmers suffer low productivity and the environment suffers degradation. Herein is outlined a complete suite of machine learning-driven systems to satisfy crop recommendation, fertilizer optimization, and yield prediction needs. The main objective is to generate intelligent, data-driven recommendations based on historical crop data, soil properties, weather data, and crop measurements so that farmers may use these data to make the best possible decisions. Random Forest models are utilized to enhance the precision of recommendations, achieving an accuracy of 62.67% for crop and fertilizer recommendation and 98.6% for yield prediction. By giving recommendations based on data and practice, this study hopes to revolutionize traditional agricultural methods and hence improve the farmer's living standards, create employment for others, and push the economy ahead in rural areas, visualizing sustainable agricultural development.