This document discusses neural net field aware factorization machines (NFFM) for predicting digital behaviors. It begins by introducing the presenters and their backgrounds. It then outlines the problem of predicting conversion ratios (CVR) and video completion rates (VCR), and why these are important. Existing models like logistic regression and tree-based models are described along with their limitations. The document then motivates the use of neural nets to find higher-order interactions and details the challenges involved. It proceeds to introduce factorization machines (FM) and field aware FM (FFM) as models, and then introduces deep neural net models like DeepFM, NFM, DeepFFM, and the proposed NFFM model. Results show NFF