What are the key considerations for deploying a machine learning solution?

Deploying a machine-learning modelling platform into a live financial crime environment is a complex task and [depending on your objectives] requires a considerable amount of planning and experience. For a large financial organisation with millions of customers and transactions per day, this could be a multi-year endeavour.<br/><br/>

In the context of fraud detection systems this requires fast response times and scaleability, as well as the ability to rapidly promote and demote new models in the new environment.<br/><br/>

The motivation in building a platform also needs to consider the requirement for model governance and demonstrate that models are constantly being aligned with the business model.