The challenge: Our customer wanted to reduce potential fraud from cash transactions among their 40,000 drivers in Malaysia.
The solution: We built a risk scoring framework that incorporated a diverse range of data, including vehicle, financial and app data, as well as geolocation and demographic information.
Technologies: Python, deep learning, Spark, supervised machine learning.
Results: Our fraud propensity predictor achieved 85% accuracy and the customer described our work as ‘transformational’.