Hilla Kalo Peled

In an era of AI explosion, it has long been understood that every organization has to build AI applications and invest resources in data science teams, but many DS departments still struggle to translate the value and contribution they bring in dollar value. This is due to the complexity of the solutions and the lack of interest from business stakeholders in scientific measurements such as recall and precision. In this talk, I will introduce the algorithms we use for accurate fraud detection, which are the essence of our organization, and how we transformed the conversation from a “nice to have” department that maintains business’ core technology into a leading department in our company that owns one of its top KPIs reported to the street. I will share our visionary ML platform, describe the pipeline for our models, and introduce our multiple measurements that are communicated to the CEO on a weekly basis.