I am an AI Red Teamer and Automation Architect operating at the critical intersection where generative AI innovation meets offensive security. I don't just build agents I systematically identify logical breaking points, data flow leakages, and orchestration vulnerabilities before they manifest as business critical incidents.

I specialize in comprehensive AI security audits. My approach begins with a forensic dive into existing business processes to pinpoint attack surfaces, followed by the development of bespoke automation primarily in Julia, Bash, JavaScript for adversarial threats.

Case Studies

• OpSyncAI Red Team Report - Technical analysis of a critical multi-agent orchestrator vulnerability chain (CVSS 9.3) that bypasses intent verification through persona adoption attacks, enabling unauthorized extraction of CRM schemas, weaponization of content engines with stealth payloads, and persistent brand poisoning via axiomatic truth injection.

• PrismGPT Variable Substitution Jailbreak - Technical dissection of a critical prompt injection vulnerability (CVSS 9.8) that bypasses AI alignment guardrails through semantic obfuscation and token smuggling techniques, demonstrating complete safety protocol deactivation via variable re-mapping attacks.

Software & Security Tooling

• ZeroSilenceAI (Whitepaper) - In-progress Cyber Reasoning System: an autonomous agentic security scanner that hunts vulnerabilities in AI agent frameworks before attackers do. Built on Bun with real Tree-sitter parsing, a No-Silent-Failure error taxonomy, and a threshold-gating decision engine.
• InjectionVault - Live informational page for an AI Red Teaming and Penetration Testing service — methodology for stress-testing LLM integrations and AI infrastructure.
• Julia Learning - A dedicated learning resource and documentation page exploring high-performance numerical analysis and automation research using the Julia language.

Open Source

• vortexVote (Whitepaper) - Dependency-free JS/Bun threshold-gating decision engine with Fast/Middle/Slow evaluation paths, built on No Silent Failure principles.
• ShadowClone (Module Doc) - Hardened JS module framework for resilient automation — scoped resource management, rate/duration/concurrency defense, and a No-Silent-Failure error taxonomy.
• AutoGenStudio-Custom-Agents - Custom agents for AutoGen Studio with specialized tools and configurations for enhanced human-in-loop conversations.
• AnythingLLM-Custom-Agents - A collection of custom AnythingLLM agents designed for automation and knowledge retrieval workflows.
• Windows File Cleaner - Batch script for Windows 11 to automate the removal of temporary files and enhance system performance securely.

I follow an evidence-first methodology for all security engagements: TRACE → BREACH → IMPACT → PROOF.

• Adversarial Testing for LLM Systems: Black-box audits of AI agents and multi-agent systems. I don't just find vulnerabilities; I demonstrate business impact through reproducible attack chains.
• Prompt Injection & RAG Risk Analysis: Forensic analysis of Retrieval-Augmented Generation pipelines, mapping trust boundaries where data flows break down.
• Agent Architecture & Trust Boundaries: Identifying handoff failures, tool authorization bypasses, and escalation pathways in autonomous agentic workflows.
• Bespoke Automation & Hardening: Crafting tailored security frameworks in Julia and Python that provide mathematical precision and high-performance probes for logic validation.
• Cloud & DevSecOps: Architecting secure, scalable cloud solutions (AWS, Azure, GCP) using infrastructure-as-code (Terraform/Ansible) to ensure automation scales without expanding the attack surface.

Agentic Security Audits (AI Red Teaming) - Learn More

• Black-box security audits for AI agents and multi-agent systems
• Evidence-first findings with reproducible PoCs (TRACE → BREACH → IMPACT → PROOF)
• No code access or prompt sharing required
• Trust-boundary mapping and OWASP LLM Top 10 alignment
• Accelerated and deep engagement packages available

1. Executive Summary: High-level risk narrative and business impact analysis.
2. Vulnerability Manifest: Technical deep-dives with reproducible PoCs (Evidence of Breach).
3. Trust-Boundary Map: Visualizing the weakest points in your agentic orchestration.
4. Remediation Roadmap: Specific code snippets and architectural changes to close the loop.
5. The Julia Advantage: High-speed security probes tailored to your specific RAG pipeline logic.

• Schedule a Call - Book a consultation to discuss your audit needs
• Email - Direct contact for inquiries

• B.S. Cybersecurity, Western Governors University
• CompTIA Security+ (2023) - Foundational cybersecurity knowledge and skills.
• CompTIA Network+ (2023) - Expertise in networking concepts and infrastructure.
• CompTIA A+ (2022) - Proficiency in IT hardware and software technologies.