Dr. Anna Bertiger is a Machine Learning Researcher at Sublime, where she uses math to find villains. Previously, she was a Principal Applied Scientist at Microsoft Security and a postdoctoral fellow in the Faculty of Mathematics at the University of Waterloo. Anna enjoys shutting down email threats with security expertise, advanced math, machine learning, and open collaboration.
Contributions
Machine learning
Scaling beyond labels: A new way to evaluate agentic learning harnesses for security
September 24, 2026
Attack spotlight
Prompt injection attacks don't look like what you’re seeing in social media and headlines
May 5, 2026
Machine learning
More than “plausible nonsense”: A rigorous eval for ADÉ, our security coding agent