Operational AI. Network Architecture. Distributed systems.
I build them, harden them, and reason about where they run, what they can reach, and what they are allowed to do.
Every agentic system eventually confronts three architectural decisions: the function, the placement, and the authority. Service decomposition, topology, access control. Old questions, but they land differently now. Function is chosen at runtime. Placement can mean the edge, where connectivity is intermittent. Authority is exercised at machine speed, often faster than a human can manage it.
The Inference Control Plane
Part One: the placement contract. Where inference runs determines what it can know, what it can do, and whether it can be trusted.
Download PDF ↓ How it fits the series →From SBOM to Decision BOM
How inventories create false confidence in AI decisions. A complete record shows how a decision was made — not whether the evidence was enough to justify it. The companion to Evidence Debt.
Download PDF ↓ How it fits the series →Evidence Debt
“The system was designed to operate, not to prove what it did.” Evidence debt compounds quietly until someone asks a question the system was never built to answer.
Download PDF ↓ How it fits the series →AWS. Principal Technical Business Developer for Mission Networking.
Worldwide Technical Leader for the National Security, Defense and Public Safety technical feedback community at AWS.
Before that, Principal Security Solutions Architect, then led the Enterprise Support team for Defense and National Security.
Before AWS, enterprise IT and distributed computing architecture.
Signals intelligence, Marine Corps. Ran a network operations center in Hawaii.
MBA, Boston University · BS Information Technology, UMass Lowell · Various AWS and other certs · Mentors veterans into cloud security