I help product teams build their own AI delivery capability.
I embed with your engineering org, build the workflows and contracts that make AI output production-ready, and leave when the team can run it themselves. Currently across 50+ teams, 200+ engineers, and 3 platforms.
The gap between pilot and production
AI pilots go well — small team, contained scope, visible results. Then it hits fifty teams, three platforms, and no shared contracts for what AI output should even look like.
Things stall not because the tools fail, but because the organization was never set up for this. I work in that gap.
How I work
The model is straightforward: I join your team for the time it takes to build the delivery system together — and step back when they can run it without me.
I work across the full range — reframing the problem with executive leadership, then building the infrastructure with your solution architects and engineering leads. Contracts, validation gates, workflows, tooling. Technical enough to do the work, not just describe it.
What I bring
AI-assisted product delivery
From Figma intent to production across web, iOS, and Android, with AI handling the mechanical translation so engineers can focus on what's actually ambiguous.
Design systems as AI infrastructure
Design tokens, component contracts, and compositions become machine-readable — so AI generates UI that actually fits the system instead of approximating it.
Agentic workspace design
Multi-agent orchestration, context packaging, and repeatable workflows for coding agents — tested across Copilot, Claude, and OpenCode.
Enterprise scale
My current work runs across 50+ teams, 200+ engineers, and 3 platforms. The problems I solve are organizational, not just technical.
Writing
I write about AI-assisted delivery, design systems, and what it's like to make these things work inside a large organization.
Read on Substack ↗