Where Watson meets Sherlock

Applying Artificial intelligence in the fraud and compliance area. How to catch a crook using AI.
When you look at it at first sight it looks obvious. Only we have a very screwed dataset, and the targets are ever moving.
What architecture to use, and how to start building and running an AI that not only catches the bad guys and at the same time is versatile and flexible while being explainable.

Overview on applicable aI for Fraud and compliance
Techniques we used since the 2008 onward
Some ML research questions we have nowadays