This white paper outlines some of the most important considerations for managing risk in machine learning models to create more accurate and compliant algorithms. Key recommendations include focusing on the quality of input data as well as implementing techniques to reduce and expose bias.