Language
Python, Java, Shell
Tool Type
Algorithm
License
Apache License, Version 2.0
Version
7338a4b
IBM
AMLSim is a multi-agent simulator designed for money laundering detection. It generates synthetic data that allows researchers to develop and test new algorithms in a unified data environment. This tool facilitates the creation of complex money laundering scenarios, providing a valuable resource to improve detection and prevention strategies in the public and private sector.
The tool solves the problem of lack of unified data and accurate simulations for anti-money laundering research. It provides a multi-agent simulator and synthetic data, allowing researchers to design and test new algorithms efficiently and consistently.
The tool functions as a multi-agent anti-money laundering (AML) simulator, generating synthetic data that allows researchers to design and implement new algorithms. It facilitates experimentation and the development of innovative solutions in a unified data environment.
AMLSim produces synthetic financial transactions in CSV files described by an open schema.json; parameters and outputs are also managed in JSON. This allows graph engines, such as Pandas, or Spark, to load the data structure and compare algorithms with reproducible accuracy without proprietary tools.
Connect with the Development Code team and discover how our carefully curated open source tools can support your institution in Latin America and the Caribbean. Contact us to explore solutions, resolve implementation issues, share reuse successes or present a new tool. Write to [email protected]
Sample CSV file showing simulated financial transaction data for money laundering analysis.
CSV file showing suspicious transaction alerts, including account details and case escalation status.
Preview of accounts.csv showing account details, initial balances, and fraud or suspicious flags.
Visit the first Security and Justice Evidence Bank available in Spanish and Portuguese, with systematized evidence of nearly 100 types of solutions and 700 cases evaluated to prevent crime and violence.