This project concentrates on developing a machine learning ensemble model to detect credit card fraud. By integrating techniques like Neural Networks, Random Forest and Gradient Boosting, the system scope to enhance fraud detection accuracy and reduce false positives. This project involves data preprocessing, model training, feature engineering and evaluating using metrics like F1-score, precision and recall. The final results offer a robust, scalable solution for real-time fraud detection, helping financial organization and minimize losses to enhance customer trust. This project integrates by ensemble methods to provide importance in various algorithms to detect complex fraud patterns in improper transacting data.