AlmaU students took part in the international HackAlem AI hackathon in Astana. One of the university’s teams developed Money Graph, an application for analysing money transfer networks and identifying potentially suspicious connections.
The event took place on 23 September. According to a post published by SDTE AlmaU, the hackathon brought together approximately 2,500 developers, students and young professionals from 21 countries and 129 universities, and was officially recognised by Guinness World Records as the world’s largest AI hackathon.
Participants included AlmaU students enrolled in the Information Systems, Data Science, Software Engineering and Product Management programmes. Over five hours, teams used modern artificial intelligence tools to develop solutions to practical challenges faced by businesses and government.
The team of Ilya Kim, Rodion Kekhaev and Yerassyl Zekebayev focused on analysing money transfers. Their Money Graph application makes it possible to examine individual transactions in the context of the entire network, showing the direction of fund flows, participants’ roles and the connections between them.
The solution processes source files in Parquet format, calculates analytical metrics, identifies participants’ roles using explainable rules and groups them into clusters. Based on the processing results, the application produces a prioritised list of nodes with explanations, helping analysts determine which areas warrant further investigation.
Ilya Kim was responsible for developing the user interface with Next.js and the interactive graph with Cytoscape.js. He implemented features for displaying transfer directions and participants’ roles, searching for nodes and viewing their connections.
Rodion Kekhaev developed the analytics component in Python, including metric calculations, explainable rules for identifying roles, clustering and ranking. He also prepared the methodology for analysis and for verifying the accuracy of the results.
Yerassyl Zekebayev developed the backend and API with NestJS, enabling data uploads and validation, execution of the Python analytics workflow and delivery of the results to the interface. The development process followed test-driven development (TDD).
A key principle of Money Graph is the explainability of its findings. The application flags potentially suspicious indicators for further examination; these indicators do not, on their own, constitute evidence of wrongdoing.
Participation in HackAlem AI enabled AlmaU students to apply their knowledge to a practical challenge and gain experience in collaborative development under time constraints. In Money Graph, programming, analytics and visualisation came together in a tool that helps reveal the connections between individual financial transactions.