My work at Merritt Collective connects offensive security practice, frontier-model cybersecurity evaluation, and research on how people learn cybersecurity.
Through Merritt Collective, I conduct and review adversarial cybersecurity evaluations of generative AI systems. I design extended multi-turn trajectories and assess technical correctness, reasoning quality, trajectory progression, demonstrated impact, hallucinations, and unsafe behavior.
My work spans software exploitation, enterprise environments, critical infrastructure, industrial controls, and space systems. My interests include agentic AI security, evaluation and benchmark design, and cyber harm and risk frameworks.
A Python research-collection project that automates daily collection, security-focused filtering, categorization, reporting, and issue generation across agent security, prompt injection, retrieval-augmented generation, and adversarial machine learning.
Explore AI Security Radar on GitHub.
MELON Masked Re-Execution — Rocky Mountain Information Security Conference, 2025.
Quantifying LLM Jailbreak Risk: MITRE ATLAS Mapping — AFA Warfare Symposium, 2025.
Cognitive Diversity in Cybersecurity Education: A Participatory Action Research Study
Doctor of Computer Science, Colorado Technical University
My doctoral study examines the learning experiences of neurodiverse cybersecurity undergraduates. It uses participatory action research, Universal Design for Learning, and neurodiversity frameworks to explore technical learning with different forms of support, including an AI mentor.
ORCID: 0009-0003-4416-8884 · OSF Preprints · Beyond Merit
For research collaboration or speaking inquiries, contact me through LinkedIn or Merritt Collective.