CSE498, Collaborative Design, Spring 2019
Computer Science and Engineering
Michigan State University

Proofpoint is a leading cybersecurity firm which provides comprehensive, cloud-based security that protects organizations from malware threats.

Every day, Proofpoint stops billions of attacks on email, mobile apps and social media accounts. This massive volume of attacks requires an efficient method for detecting malware.

Our Defeating Malware Payload Obfuscation platform provides a faster and more efficient way to determine whether incoming files are benign or malicious. Our system utilizes a machine learning approach to detect and neutralize malware payloads.

Among other things, our platform detects so-called obfuscated malware, in which an attacker hides malware in a seemingly innocent document, such as a photo. Such a diagnostic process can be difficult and expensive. By handling different file types separately, our machine learning algorithm quickly and accurately classifies a wide range of malware files.

Our platform includes a companion web dashboard that displays basic system information, including a system health information page, and pages that examine details of the classification of an individual file and allow the user to submit files manually to be analyzed.

Our backend platform uses a Python controller to extract metadata from different file types and feeds that information into our machine learning algorithm running Keras, Tensorflow and scikit-learn to make a classification. Our web dashboard uses Flask for the backend, and Bootstrap, HTML and JavaScript for the frontend.