Skip to content
View rexcoleman's full-sized avatar
1 follower · 0 following

Highlights

  • Pro

Block or report rexcoleman

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rexcoleman/README.md

Rex Coleman

Securing AI From The Architecture Up

Founder & CEO of Singularity Cybersecurity. Building and attacking AI security systems at every layer of the stack — then publishing the methodology so others can too.

Current Work

  • AI Security Research — systematic red-teaming across RL agents, LLM agents, model backdoors, adversarial ML, vulnerability prioritization, and post-quantum cryptography
  • govML — open-source ML governance framework (50+ templates, 20+ generators). Contract-driven reproducibility for ML experiments.
  • Singularity Cybersecurity — AI security research lab. Building AgentArmor, SkillVet, and RedClaw [all HYPOTHESIZED — in design phase]

Background

15 years cybersecurity — data analytics and enterprise sales at FireEye/Mandiant, then Fidelis. Now building the security layer for AI systems.

MSCS Georgia Tech (Machine Learning) · CISSP · CFA Charterholder

Links

Pinned Loading

  1. adversarial-ids-ml adversarial-ids-ml Public archive

    Adversarial ML on network IDS: feature controllability constraints reduce attack success 35% and enable architectural defenses that outperform adversarial training. CICIDS2017, 5-seed evaluation, g…

    Python

  2. vuln-prioritization-ml vuln-prioritization-ml Public

    ML-driven vulnerability prioritization: predicting which CVEs get exploited. 4 research questions, SHAP explainability, 11 architectural decision records. govML-governed.

    Python

  3. agent-redteam-framework agent-redteam-framework Public archive

    Open-source security testing for LLM-based agents. 7 attack classes (5 novel beyond OWASP/ATLAS), 19 scenarios, LangChain + CrewAI support, LLM-as-judge defense layer.

    Python

  4. financial-anomaly-detection financial-anomaly-detection Public

    CFA x ML fraud detection: XGBoost AUC 0.987, 8 CFA-informed features in top 20 SHAP, 81% adversary-resistant floor. 5th domain controllability analysis. govML-governed.

    Python

  5. pqc-migration-analyzer pqc-migration-analyzer Public

    Post-quantum crypto migration scanner: detects quantum-vulnerable primitives, ML priority scoring (+14pp vs rules), NIST PQC mapping. 21K crypto CVEs analyzed. govML-governed.

    Python