I work with an amazing team of engineers and security and AI experts on measuring and mitigating security risks in Llama models, ensuring Llama gets deployed securely within Meta's apps, and open sourcing AI security tools and findings

I led scientific research @ Sophos, focused on the intersection of cybersecurity, machine learning, data visualization, and large scale scientific computing infrastructure

I led machine learning / deep learning cybersecurity research at Sophos.

I led security data science research efforts at Invincea before its acquisition by Sophos. Particular focus on inventing, evaluating and deploying deep learning detection models in support of a next-gen endpoint security solution.

I provided technical leadership on programs exploring the application of machine learning and data visualization methods to cybersecurity problems. I also contributed to proposals to scientific funding agencies, wrote papers for academic computer security conferences, and helped explore transition opportunities for our research accomplishments.For an article on a talk I gave at Blackhat 2015 surveying data science work at Invincea see:http://www.darkreading.com/vulnerabilities---threats/black-hat-data-science-machine-learning-and-you!/a/d-id/1321771For a popular article on our work, see:http://www.newscientist.com/article/mg21728985.300-mapping-malwares-genome-to-fight-future-attacks.htmlAbstract describing our CrowdSource project:http://www.blackhat.com/us-13/briefings.html#Saxe

I invented, implemented, and documented cyber-security tool prototypes. I did packet / application level security analysis as well as data visualization, data mining, and UI design.

Projects: Co-led a real-time honeypot forensic analysis project. Also, did security data visualization, exploit and rootkit research on Windows and Linux platforms.