The environmental impact of data centers has become a major concern. In their Q1 2021 report, Intel estimated that there are around 100 million servers deployed globally. In this paper we provide an overview of Interact, a machine learning tool that helps data center operators easily analyze, manage, and optimize the energy, cost, and Carbon footprint of their current servers and infrastructure. We discuss in detail the machine learning algorithms and the energy models used. We also demonstrate the key features of the tool and how it can help data centers become more sustainable by reducing server inefficiencies.