This research project focuses on mitigating the problem of spam email attacks by exploiting the spammer network's human resources. The goal is to create a context-aware language-based computer program that understands and is able to respond to human natural language. More specifically, the program will utilize neural network and other machine learning techniques to extract nonlinear patterns found by analyzing data containing spam conversations. The goal is to deploy a system that can hold conversations with spammers at scale, exploiting spammers' time and mitigating the problem of spam traffic on the web.