The increased human population increases the demand for food. Traditional farming leads to inefficiencies and difficulty in fertilizer usage, crop selection, and insect detection. This research project eliminates all these problems by developing an advanced farming web application to evaluate crop production efficiency. This research evaluates the soil nutrients needed by different plants and thereby generates a recommendation system to recommend the most suitable crop based on sensor values, thus reducing risk, nutritional imbalance and environmental pollution. It consists of an NPK sensor with a combination of machine learning models to monitor soil health and increase yields, reduce costs and match fertilizer supply with demand. It also helps to analyze various insects and provides descriptions of insects and recommends solutions to those insects in Nepali language helping farmers. The integration of ML and DL models such as random forest for fertilizer prediction, light GBM (Gradient Boosting Machine) for crop prediction and Conv2D for the classification of insects will help to maximize the production yield.