https://github.com/adrn-mm/transactional-network-analysis
In the modern financial landscape, trillions of transactions occur every single day. While tabular data (rows and columns) is excellent for storing these records, it fundamentally fails to capture the “big picture”—the intricate relationships and connections between different entities.
The Transactional Network Analysis project was developed to bridge this gap. It is a highly interactive, AI-driven web application designed to explore, analyze, and visualize complex webs of financial transactions, whether they occur between individuals (personal) or large-scale institutions.
By mapping the flow of funds visually and augmenting it with machine learning, analysts can rapidly detect anomalies, understand the hierarchy of fund distribution, and investigate potential money laundering rings or fraudulent activities that would otherwise remain hidden in millions of rows of logs. As a testament to its robust architecture, analytical depth, and immediate business value, this project was awarded 1st Place at the Cloudera AI Hackathon 2025.