30 years of building, breaking, and fixing technology — from Navy field medicine and datacenter operations to SaaS enterprise support leadership. I think in systems, document obsessively, and build things that last.
Thirty years of watching technology solve the wrong problems — and occasionally the right ones — generates opinions. This is where I put the ones worth writing down.
The first generation of large language models proved the concept works. They also built on a structurally flawed foundation — epistemic flattening, where peer-reviewed research and a Reddit thread carry identical weight in the training pipeline. A market correction is coming. HEAL is a design specification for the architecture that comes after it.
The framework organizes AI knowledge into four stratified tiers with distinct governance rules, confidence metadata, and modification standards. The result: an AI that knows what it knows, knows what it thinks, and knows what it prefers — and is architecturally required to communicate the difference.
The professional life pays the bills. The homelab keeps the mind sharp. The fiction writing keeps everything honest.
Director-level SaaS production support with a track record of building support organizations from scratch — including 98.3% shift-left success, 42% reduction in engineering escalations, and 100% P1 SLA compliance. ITIL certified. Twenty-five years of progressive operations leadership across SaaS, enterprise IT, and logistics.
A privately operated, production-disciplined home infrastructure running dual network zones — a segregated DMZ (Apex) and a household LAN (Monarch) — with OPNsense firewall enforcement, Proxmox hypervisors, Docker containerized services, Tailscale overlay networking, Ansible automation, and GPU-accelerated local AI inference. Documented to enterprise standards.
A fantasy novel in progress. More when it's ready.
"Simple, stable, boring equals safe and secure."
The design philosophy behind the Tinmachi homelab — but it applies equally to the professional work. I don't chase novelty for its own sake. I build things intended to still be running in five years, documented well enough that someone else could operate them, and scaled to the actual problem rather than the hypothetical one.
Navy Hospital Corpsman, 1994–2002. Eight years of service that taught me what high-stakes operations actually look like — and that documentation saves lives. The IT career that followed has been one long application of that lesson.