Building a machine learning operations platform and data experimentation for Anti-money laundering detection

Abstract

With the robust development of cloud computing, the explosion of big data, and the continuous advancement of science and technology, machine learning (ML) has been and is being widely applied across various industries. The arising issue is how to increase the performance and reliability of ML applications and minimize the go-tomarket time in the process of building and applying AI/ML. As a result, a series of standards for operating and deploying ML projects have gradually been formed, opening up a completely new direction for Machine Learning Engineers and Data Scientists, known as Machine Learning Operations or MLOps. This aims to accelerate the process of bringing ML products to production more quickly and efficiently. In this thesis, I have integrated the Machine Learning Operations (MLOps) into Anti-Money Laundering (AML) detection enhances the ability of financial institutions to identify and prevent money laundering activities. This approach involves the use of machine learning models and automated processes to improve the efficiency and accuracy of AML systems.