Charleston, South Carolina Metropolitan Area
work at the intersection of cybersecurity, machine learning, and AI security, specializing in securing generative and agentic systems against adversarial behavior. My background in cybersecurity gives me a deep understanding of identity, authorization, data‑flow, and attack‑surface risks, while my machine learning experience allows me to diagnose model behavior, failure modes, and reasoning vulnerabilities with precision. I build and run adversarial evaluations, diagnostic probes, and stress‑testing frameworks to uncover weaknesses in LLMs and autonomous agents. My hands-on work spans AWS Bedrock, Azure AI, and Google Vertex, where I design taxonomies, harnesses, and evaluation pipelines that expose model vulnerabilities and guide remediation. I also work extensively with agentic frameworks such as LangChain, CrewAI, and Hugging Face Transformers to orchestrate secure autonomous workflows and simulate adversarial behavior at scale. My approach blends AI-native red teaming, classical cybersecurity principles, and practical machine learning knowledge to identify risks that traditional testing misses.
I operate as an AI Security and Cybersecurity Consultant, helping organizations secure generative systems, agentic workflows, and cloud‑native AI deployments. My work blends AI security, cybersecurity engineering, and data science to identify vulnerabilities, harden autonomous systems, and strengthen operational integrity. I work across the intersection of AI security, cyber engineering, machine learning, and data‑driven diagnostics, enabling clients to understand and mitigate risks that traditional security testing cannot detect. Using agentic frameworks such as LangChain, CrewAI, and Autogen, I: • Design and orchestrate secure multi‑agent systems • Simulate adversarial behavior and red‑team scenarios • Fine‑tune agents for task‑specific performance • Integrate data science workflows to analyze model behavior, detect anomalies, and validate outputs • Apply cybersecurity engineering (IAM, authorization, data‑flow analysis, threat modeling, attack‑surface reduction) to secure AI pipelines end‑to‑end I operationalize industry‑standard AI‑security tooling, including: • PyRIT for structured LLM risk mapping • Promptfoo for prompt harness benchmarking • Garak for automated vulnerability scanning I build reproducible payloads targeting: • Schema confusion • Privilege erosion • Agent‑level drift • Data‑driven failure patterns and ML‑based misalignment signals My consulting practice combines AI‑native red teaming, cybersecurity engineering, and data‑science‑driven diagnostics to expose and remediate vulnerabilities across modern AI systems.
Delivered security engineering, cloud hardening, and threat‑driven assessments across diverse client environments. This period became the foundation for my transition into AI security, where I began integrating data science and machine learning workflows into security operations. Applied ML‑driven analysis to detect anomalies, evaluate model behavior, and validate system outputs. Used agentic frameworks to build early multi‑agent prototypes, orchestrate task‑specific agents, and fine‑tune agent behavior for reliability and consistency. Strengthened cloud security posture through hands‑on assessments, vulnerability scanning, and incident response. This role evolved into a hybrid practice combining cybersecurity, cloud engineering, and emerging AI security techniques — ultimately leading to my current specialization in securing generative and agentic systems.
Monitored, diagnosed, and resolved network issues across 500+ client systems, maintaining 99.9% uptime and ensuring stable, high‑performance connectivity across distributed environments. Led critical outage response, reducing resolution time by 30% through structured troubleshooting and escalation workflows. Analyzed network performance data to identify patterns, optimize configurations, and implement preventative measures — early work that shaped my later transition into data science and machine learning–driven analysis. Strengthened network security posture through audits, protocol hardening, and proactive vulnerability reduction.
Maintained 99.9% network uptime across large‑scale broadcast and distribution systems by applying advanced monitoring, diagnostics, and structured troubleshooting. Resolved 200+ network incidents with an average 15‑minute response time, minimizing downtime and ensuring SLA compliance. Planned and executed 15+ network upgrades annually, integrating new technologies without service interruption. Continuously monitored 200+ systems — servers, network infrastructure, applications, databases, and data centers — to proactively identify issues and reduce incident recurrence by 25%. Analyzed weekly performance metrics and incident patterns, producing data‑driven recommendations that improved network efficiency by 15% and reliability by 10%. This early exposure to large‑scale systems and operational telemetry became the foundation for my later work in data science, machine learning, and AI‑driven diagnostics.