AI Product Manager

Building AI systems that
work in production.

Not demos. Eight years shipping at the intersection of LLMs, agents, and enterprise scale — from a $20M digital services business at a Fortune 500 to a 54-agent autonomous OS.

View Work National Grid Case Study →
Core Capabilities
⚙️
Agentic Systems
Shipped a 54-agent production OS running on Claude with CEO/worker tiers and durable orchestration.
🛡️
Safety Engineering
ethos-gate: application-layer safety framework for autonomous agents — hooks, credential policy, blast-radius doctrine.
🏢
Enterprise AI
Scaled an AI product from $2M to $20M annualized revenue at National Grid (Fortune 500).
📊
Infrastructure
Cost optimization, model tiering, fleet orchestration, and account routing at production scale.
🧪
Evaluation
Built production eval frameworks tied directly to business outcomes, not benchmark scores.
⚖️
AI Governance
Autonomy ladders, human-in-the-loop design, and lethal-trifecta audit patterns for agentic systems.
Work
clawd-agent-os
OSS · MIT
Production framework for orchestrating hierarchical Claude agent fleets

Extracted from a live 54-agent system. CEO/worker tiers, durable inter-agent bus, fleet task drainer, and blast-radius-aware model routing. Covers the orchestration layer — how to design, spawn, and run a fleet of specialized agents without them stepping on each other or running away on cost.

GitHub → Architecture docs →
ethos-gate
OSS · MIT
Application-layer safety framework for agentic Claude systems

PreToolUse/Stop hooks for Claude Code, three-tier credential scrape policy, irreversibility guards, and the "lethal trifecta" audit (data access + untrusted input + outbound channel = leak vector). 62 tests. Ships with an autonomy ladder as Python dataclasses and a real blast-radius doctrine — not a list of rules.

GitHub → Docs →
National Grid
Case Study
Enterprise AI operating model — $2M → $20M annualized revenue

Built and ran the fixed-price digital services business at National Grid from $2M to $20M. Expanded a cloud portfolio from 0 to 500+ services. Launched an internal GenAI platform with 3,000 monthly active users. Built an AI dynamic pricing engine that improved margins 22%. The playbook for deploying AI at Fortune 500 scale — not pilot scale.

Case Study →
About
Previously
Principal PM, National Grid (Fortune 500)
 
Co-founder / CPO, Dominium Finance
 
Senior Full-Stack Developer → PM
Currently
AI PM & MBA Candidate
 
BYU Marriott School of Business
 
Strategy & Product Management, 2025–2027