Cyber Defense Through Engineering and Analytics
Cyber Defense Through Engineering and Analytics
I bring an unusual combination to security engineering: formal training in computational mathematics and applied statistics, hands-on detection engineering across Splunk and Microsoft Sentinel, and active dissertation research in adversarial ML detection in autonomous systems. My work sits at the intersection of rigorous probabilistic modeling and operational security — building detection architectures that reason correctly under adversarial conditions, not just normal ones.
Ryan Beavers
Cybersecurity & Systems Engineer
Threat Detection | Network Defense | Secure System Design
I build detection systems that hold under adversarial conditions — combining rigorous probabilistic foundations with hands-on security engineering across SIEM, anomaly detection, and ML-based threat analysis.
Email:
Location:
Portland, Oregon
Core Capabilities:
Adversarial ML detection · Bayesian/Kalman anomaly architecture · Splunk SIEM · Microsoft Sentinel · KQL · MITRE ATT&CK · Python · Active Directory hardening · Threat modeling (STRIDE/DREAD) · D.Eng. Cybersecurity Analytics, GWU · M.S. Applied Statistics, OU · Intel SSG Engineering Excellence Award
EXPERIENCE
2024-Present
Graduate Researcher
Graduate Researcher — Cybersecurity Analytics GWU / University of Oklahoma | 2024–Present
Conducting applied research in adversarial ML detection, anomaly detection architecture, and probabilistic security modeling. Designing a layered Bayesian/Kalman filter architecture (Bronze/Silver/Gold trust tiers) for detecting adversarial manipulation of sensor inputs in autonomous systems — proof-of-concept: GPS spoofing in drone platforms. Detection coverage validated against MITRE ATT&CK tactics. Producing dissertation documentation bridging academic research and operational security engineering.
2021-2022
Teaching & Technical Communication
Portland State University
I helped students communicate complex analytical concepts clearly and accurately, improving documentation and structured reasoning—skills directly applicable to cybersecurity reporting and collaboration.
2008-2010
Software Engineer
Intel Corporation
I engineered real-time telemetry and system simulation tools that accelerated performance feedback and reliability insights for new hardware. I collaborated with cross-functional teams to optimize system efficiency and earned multiple recognition awards for my contributions.
Doctor of Engineering
Beginning January 2026
The George Washington University
Cybersecurity Analytics
2024-2025
Master of Science
UNIVERSITY of Oklahoma
Applied Statistics
2025
Certificate
Fullstack Academy
Cybersecurity
2019-2021
Master of Arts
Johns Hopkins University
Liberal Arts
2004-2008
Bachelor of Science
Arizona State University
Computational Mathematics
PROJECTS
ETHICS & CRITICAL ANALYSIS
This essay examines how generative AI tools shape our understanding of human values — and what happens when ethical context is missing. By analyzing AI-generated definitions of leadership, I explore the risks of “bias by omission,” the illusion of neutrality in algorithmic systems, and the responsibility humans retain for moral judgment in technological environments.
Automated hiring promises efficiency — but without ethical safeguards, it can quietly reinforce discrimination instead of eliminating it. This analysis examines Amazon’s failed AI recruiting tool and shows how biased training data, opaque decision logic, and misplaced trust in automation can harm real people seeking real jobs. It calls for transparency, fairness auditing, and human accountability in the use of machine-learning systems that influence life-changing opportunities.
Every analyst has a go-to way of making sense of the world. For me, data behaves like language—full of structure, rhythm, and meaning beneath the surface. This narrative explores how my background in the humanities and applied statistics shaped a unique analytical mindset: treating logs as stories, anomalies as “glitches,” and cybersecurity as a communication problem between humans and machines. It also reflects critically on the limits of any single method—and the importance of expanding the toolbelt.