Five years bringing FAR/DFARS discipline to Department of Defense programs — and now bringing the tools that modernize it. I independently architected and ran an autonomous multi-agent AI operation end to end: six specialized agents, hard-stop cost controls, zero-trust security, all on scheduled cloud workers.
The through-line isn't a career pivot — it's the same skill applied to a new medium. Acquisition taught me to scope requirements, govern cost, and hold a program accountable to delivery. Applied AI is where I now do it, at machine speed.
Contract lifecycle management, cost/price analysis, requirements gathering, and ERP-based financial tracking across active DoD programs — governed by FAR/DFARS and delivered in Agile/SAFe cadence.
Agent orchestration, workflow automation and CI/CD, cost-governance controls, secure cloud pipelines — the same rigor, now building the systems instead of only administering them.
A self-directed project: a streetwear brand and open-world game operated by a crew of autonomous AI agents on free, scheduled cloud compute. Each agent owns a role, pulls its own work from a queue, reports back, and asks for my approval before anything ships. I set the roadmap; they do the work.
The big brain. Runs market and competitive research, breaks goals into tasks, assigns work to the crew, and routes my orders to the right agent — all queued for my approval.
Owns the go-to-market side: campaign planning, audience analysis, and the morning brief — turning raw activity into decisions I can act on before coffee.
Generative product design — apparel mockups and concept art produced on demand, staged in a review gallery so nothing goes live without a human yes.
The studio: launch videos and 3D assets, phased with per-item cost breakdowns so every dollar of production spend is scoped and approved before it's spent.
The engineer. Runs nightly build shifts, ships features, does its own QA and cleanup, and leads a specialist sub-crew for the game engine work.
Automated daily security sweep — scans for vulnerabilities and anomalies, watches the perimeter, and flags anything that needs a human eye.
Per-service monthly/weekly budget caps with hard-stop enforcement, live spend tracking, and projection dashboards — acquisition cost discipline applied to real operating spend, toggleable and editable on the fly.
Cloudflare WAF, SSO login wall (Zero-Trust Access), managed DNS, and secret management — plus an automated daily security-audit agent watching the perimeter.
The entire crew runs on scheduled GitHub Actions workers against a PostgreSQL backend with row-level security — event-driven automation on free compute, no server to babysit.
A self-hosted, MIT-licensed zero-shot TTS engine running on free runners — clone a voice from a sample, generate game dialogue, at $0/mo instead of a paid API subscription.
A command center for KPIs, a "waiting on you" approval queue, morning reports, and a live calendar — the human-in-the-loop layer that keeps me in control of every autonomous action.
A modular skill system that injects capabilities into each agent, plus an XP/leveling model that expands an agent's reasoning budget as it proves itself on real work.
Applied AI, automation, product ownership, or acquisition — with the clearance to do it where it matters.