Transaction Harbor is a web application designed to enhance the efficiency of risk analysis processes by automating the analysis of transaction data while saving time and resources. The proposed solution involves developing a system that automates the processing of transaction data, performs risk calculations, and flags potentially fraudulent transactions. By automating these processes, risk analysts can focus on investigating high-risk transactions and implementing proactive measures to mitigate fraud. Additionally, the integration of Streamlit enables the creation of interactive dashboards and reports, providing stakeholders with real-time insights into transaction trends and patterns.
This page facilitates detailed breakdowns and analysis of transaction data. Users can upload CSV files containing transaction records and specify parameters such as filenames, customer names, and high-ticket thresholds for analysis. Upon uploading, the app performs various calculations, offers functionalities such as filtering transactions based on specific criteria, identifying high-value transactions, and detecting potential instances of suspicious behavior. Moreover, users can download Excel worksheets containing comprehensive data summaries for further examination and reporting.
This page offer users the functionality to explore their transaction data through the Chart Maker tool. With parameters such as payment status, method, application, and country, users can refine their analysis, while setting start and end dates allows for focused examination over specific time frames. Four visualization types—Histogram, Box and Whiskers, Box Plot Sum, and Box Plot Count—enable users to glean insights into transaction trends, counts, total amounts, and distribution across payment types. This empowers users to make informed decisions, detect anomalies, and enhance risk management within financial institutions.
