Building practical AI tools for security, operations, and decision support. 24-year Army veteran with 5 combat deployments, cybersecurity professional, AI builder, and founder/operator working at the intersection of security, automation, and applied AI.
A live, video-inspired knowledge tree built from curated portfolio metadata. Private repository details and sensitive implementation notes stay out of the public view.
Open Public BrainTap nodes to identify them. Open Public Brain for node descriptors.
My work ties together 24 years Army leadership — including 5 combat deployments — UC Berkeley MICS cybersecurity training, AI/software projects, and a founder/operator mindset. That combination supports roles and ventures across cybersecurity, AI automation, technical leadership, consulting, and startup execution.
I build systems that are practical first: agent workflows that save time, debate simulation that supports civic reasoning, computer-vision tools that make data useful, and security controls that keep automation accountable.
The Human Rights First fellowship and Debate Bot work reflect the same through-line: applied technology should help people make better decisions.
Across AI automation, cybersecurity, and product building.
Deployed AI operations and automation platform for turning messy client requests into structured workflow briefs, risk notes, and implementation plans.
Policy-as-code framework for defining what autonomous agents may read, write, execute, or escalate before touching sensitive systems.
Public-safe agent orchestration concept for delegated coding, research synthesis, scheduled reporting, and tool-governed workflow automation.
Deployed RAG-powered political debate simulator connected to Code the Vote and Human Rights First fellowship work, focused on grounded civic reasoning.
Food recognition and macro-estimation app using image input, nutrition data, and model-assisted inference to produce useful meal estimates.
QR-based passenger song request app for rideshare drivers, designed around lightweight mobile interaction and low-friction queue management.
Security monitoring and watchdog concepts for governing autonomous tools, enforcing policy, and making AI-agent behavior auditable.
Healthcare safety prototype for reviewing medication-risk concepts with synthetic staging data and strict public-safe boundaries.
Working AI-agent systems documented in depth.
OpsPilotAI is a deployed AI operations concept built around a practical question: before automating a business workflow, can the team define the problem, pilot scope, risk posture, human-review requirements, and success measures clearly enough to act safely?
Useful AI and automation lessons were arriving as videos, but raw links are not durable knowledge. This workflow turns source material into searchable, provenance-preserving operating context without wasting tokens or publishing private internals.
The learning system produced useful JSON ledgers, Markdown demos, skills, and scripts. This case study shows the layer that turns those artifacts into an auditable implementation queue.
Simple structured events should not wake a model on a schedule. This case study shows how the YouTube learning workflow moved toward cheap event detection and reserved LLM reasoning for actual analysis.
Selected open-source work framed by the concept, public value, and implementation pattern — without private repo details.
Concept: make a brittle auth-dependent learning workflow easier to understand and maintain. The contribution documents SAPISID hashing behavior so future users can debug the YouTube-to-study-notes path without relying on hidden context.
View ProjectConcept: tighten local safety boundaries for assistant tooling. The contribution hardens localized link-scheme checks so cross-platform AI assistant behavior is more predictable, auditable, and fail-closed.
View ProjectAvailable for cybersecurity, AI automation, and technical leadership opportunities. Best entry points are LinkedIn, GitHub, and project-specific demos.