This document discusses automated decision making with big data and predictive applications. It begins by looking at how decisions are currently made, which is often based on qualitative factors rather than data-driven insights. It then examines how predictive applications work by collecting and analyzing data to build models that can make decisions and be tested and optimized. The document argues that predictive applications can help reduce risk and costs while increasing revenue by enabling trends to be estimated, classifications to be made, and events to be predicted to optimize returns. It presents the idea of having a common platform for building, running, and monitoring predictive applications using both internal and external data.