Cybersecurity • Secure Systems Engineering • Cyber-Physical Systems • Trustworthy Automation • AI Safety & Security

Founder, BattleReef Ltd. | U.S. Army Combat Veteran | IT & Cybersecurity Student | Open-Source Builder

I build systems where security, software, physical devices, data, and human decision-making intersect.

I am completing a B.S. in Information Technology and Cybersecurity at Southern New Hampshire University while developing open-source security, automation, and data systems. My work increasingly focuses on the engineering problems that arise when software decisions can produce real-world consequences: identity, authorization, observability, verification, fault tolerance, evidence, and safe failure.

Before moving into technology, I spent more than two decades serving in the United States Army, including multiple combat deployments. That experience continues to shape how I approach engineering: understand the mission, minimize unnecessary trust, verify critical assumptions, preserve evidence, plan for failure, and build systems that remain useful under imperfect conditions.

I am also the founder of BattleReef, Ltd., a marine technology company focused on large-scale automation, environmental monitoring, and research-oriented systems for aquatic environments.

I am especially interested in AI safety and security, trustworthy automation, human-AI collaboration, cyber-physical security, critical infrastructure protection, and resilient system architecture.

A secure, event-driven cyber-physical platform for marine monitoring, automation, and device control.

BattleReef is my primary engineering project and a practical testbed for questions that matter across both cyber-physical and AI-enabled systems:

• How do we authenticate machines and services rather than merely trust network location?
• How do we detect when authenticated telemetry is still wrong?
• How do we distinguish command intent from successful physical execution?
• How can independent evidence challenge a component's own report of its state?
• How do we preserve an auditable record when something fails?
• How should high-impact automated actions be bounded, observable, and recoverable?

The current platform includes FastAPI, SQLAlchemy, PostgreSQL/TimescaleDB, MQTT over TLS with X.509 device identity, role-based access control, telemetry plausibility analysis, quarantine, tamper-evident hash-chained auditing, device-health evaluation, safety watchdogs, independent physical-state verification, fault-injection tooling, React/Vite interfaces, Docker, and GitHub Actions security testing.

A core design principle is simple: a command being sent is not proof that an action occurred. BattleReef separates command intent, acknowledgment, telemetry, and independent physical evidence so contradictions can be detected instead of silently trusted.

A public coastal-conditions, environmental-data, and research platform for knowthegulf.com.

Know the Gulf combines historical beach-flag observations with current observations, weather, waves, tides, and modeling-ready datasets. The project deliberately separates official public-safety information from statistical and model-generated output, because useful predictions should not be presented as substitutes for authoritative guidance.

• SOC Log Triage — explainable, rule-based log normalization and analyst triage tooling.
• Secure Notes API — secure backend patterns involving authentication, authorization, encryption, and explicit trust boundaries.
• Security Program Starter — incident-response, risk-management, governance, and security-program resources.

AI Safety & Security
Tool authorization, model and agent trust boundaries, observability, human oversight, failure containment, adversarial testing, secure AI-enabled systems, and trustworthy human-AI collaboration.

Cyber-Physical Security
Systems in which compromised software, incorrect telemetry, stale data, unsafe automation, or device failures can create physical consequences.

Security Architecture
Identity, least privilege, defense in depth, segmentation, secure APIs, zero-trust principles, auditable control planes, and resilient infrastructure.

Trustworthy Automation
Automation that constrains authority, preserves evidence, detects contradictions, and fails safely rather than simply acting faster.

The questions that interest me most are usually trust questions:

What are we implicitly trusting?

How would we know if that trust were misplaced?

What independent evidence could challenge the system's own conclusions?

What happens when a sensor, service, user, network path, or model is confidently wrong?

I am interested in building systems that are not merely capable, but observable, controllable, auditable, resilient, and worthy of trust.

Languages: Python • JavaScript • SQL • Bash
Platforms & Tools: Linux • Docker • Git • GitHub Actions • FastAPI • PostgreSQL • TimescaleDB • MQTT • React/Vite
Security: X.509 / mTLS • RBAC • Secure APIs • Detection Engineering • Risk Management • Incident Response • Zero-Trust Design

I am extending my work in secure cyber-physical systems toward AI agent security and trustworthy autonomous systems, particularly architectures where model or agent actions are treated as untrusted proposals that must pass authorization, policy, verification, and audit controls before producing high-impact effects.

LinkedIn • ORCID • BattleReef • Know the Gulf

Build systems that can be questioned, verified, and trusted.