Started off at Sophos working with their big data streams managing front-end access (Datameer) and back-end maintencance (Hadoop). This data was product and detection data feeds. This data was used to generate analytical reports and visualizations and to program automatic workflows to keep them up to date. The research focused on product improvement, malware classification, and generating an understanding of the threat landscape. A case study by Datameer discussed more about how this data was used using their front-end system.
Later joined the new team of Data Scientists acquired from Invincea labs to assist in reasearch on the use of deep learning models to detect malicious content in areas such as portable executables, documents, web content, and more. Additionally assisted in the management of their data feed using AWS and Redshift, generated analysis on the datafeeds, and improved the data worflow. A unique role was to maintain the relationship between the new DataScience team at Sophos and the existing SophosLabs threat research teams in regards to data sharing, data needs, and workflow. Used deep learning, python, AWS, SQL, JS, and more to conduct research and evaluation on malware detcting models.
As part of the team, we were encouraged to write about machine learning and our work with the labs. Below are the pieces I contributed to while at Sophos and as a returned contractor for a short period of time: