Fraud anomaly detection is a critical method used across industries such as finance, insurance, e-commerce, and cyber security to identify and mitigate potential fraud risks. This project focuses on developing a comprehensive framework to detect fraudulent activities by leveraging techniques. The approach integrates methods like Interquartile Range (IQR) and standard deviation analysis to identify outliers, as well as K-means clustering to group transactions and highlight anomalies. Using real, anonymized transaction data from a Czech bank spanning 1993 to 1999, this research analyzes financial patterns, focusing on October 1997. Through advanced data visualization, outliers and suspicious activities are highlighted, potentially signaling fraud or money laundering. The project also employs evaluation metrics like the Silhouette score and the elbow method to optimize clustering accuracy. By implementing this data-driven approach, the proposed framework aims to enhance fraud detection, reduce false positives, and improve decision-making processes. Additionally, the integration of digital forensic investigation techniques and risk management strategies aims to reduce financial losses and minimize reputational damage, providing a robust and proactive solution for combating fraud