A diagnostic framework for design leaders building governance, capability, and operating structure for AI — before it gets defined for them.
16 years in product design · Built and deployed AI agent systems inside enterprise design teams
The pattern I keep seeing
Your VP just asked what your design team is doing with AI. You don't have a clean answer yet.
That team was using AI. People were getting faster. But when you look at what was actually happening, no one owned the quality bar for AI output, no strategy connected any of it, and leadership couldn't tell the difference between AI-assisted work and AI-driven work. That's not an adoption success story. That's a team that's exposed in the layers they haven't built yet.
The model
AI readiness is mapped as a separate axis, because adding AI to broken operations just accelerates what's already failing.
Has leadership decided what AI is actually for on this team?
When this is missing, every designer is optimizing for something different. One is using AI to speed up wireframes, another is drafting research plans, a third isn't using it at all because nobody told them they could. Protiviti's 2026 data: only about a third of organizations have reached Level 3 in AI strategy.
When someone generates a spec or a research synthesis with AI, who reviews it?
When this is missing, specs hit engineering incomplete and nobody can trace it back to the AI draft that started it. The $298K/year handoff cost per product pod that Supernova documented gets worse with AI — more artifacts moving faster with less scrutiny. Only 25% of teams enforce their own design guidelines.
Can people use AI with real judgment, or are they self-teaching through YouTube?
When this is missing, your strongest designers use AI as a shortcut and your junior designers use it as a crutch, and nobody can tell the difference in the output. 96% of designers are self-teaching AI through social media. No shared standards, no structured learning, no baseline.
Do the team's workflows and documentation actually support AI, or fight it?
When this is missing, AI runs alongside the existing system in a parallel track nobody maintains. Outputs don't connect to how work moves between teams. Only 11% of design teams have documented research processes. The tools multiply, the fragmentation compounds.
Is anyone measuring what AI is doing for work quality, or just counting who has access?
When this is missing, you can't tell whether AI is making the team's work better or just making it faster. McKinsey's 2026 data: almost two-thirds of organizations using AI haven't implemented it at scale. Only 39% can point to quantifiable impact.
Pick the layer you're weakest in. Not the one that sounds most interesting. The one that's most exposed.
Where teams get stuck
AI Chaos
High capability, no governance, no strategy.
Your team is using AI fast. Real speed gains on production work. But nobody owns the review process, there's no shared standard for what 'good enough' AI output looks like, and leadership hasn't connected AI adoption to any team-level goal. This is where AI-generated work ships without scrutiny and quality erodes invisibly.
Traditional Excellence
Strong operations, clear ownership, no AI integration.
Your workflows are solid, reviews happen, ownership is clear. But AI isn't part of any of it. You're well-run and at risk of being outpaced. Not because your work is worse, but because teams integrating AI into their operating model are building compound advantages that widen over time.
Most teams fall into one of six profiles. AI Chaos and Traditional Excellence are the two where teams don't realize they have a problem. The diagnostic tells you which profile you're in.
How I help
Free
24 questions. 5-layer scoring. Instant profile. Map where your team stands operationally and how your AI adoption is helping or creating risk. Takes about 10 minutes.
Free · Weekly
Practical frameworks for building AI-enabled design teams — delivered weekly. Real patterns from real design organizations. Published on Substack.
About
I help design leaders build a defensible operating model for AI — before their org defines the structure without them.
10+ years in product design and design operations. Most of them building and scaling design teams from the inside: governance models, skill frameworks, research operations, design system architecture, all at enterprise scale. I spent a year migrating a 20-person team's entire infrastructure from Planner to Jira and Confluence, building the connective tissue between design and engineering. That's the work this framework comes from.
What makes my approach different: I don't just design frameworks. I build the AI systems that make them run. Gemini Enterprise agents, Atlassian Rovo agents, custom Gems, deployed into real design workflows. Not slide decks.
The diagnostic maps your team across 5 operational layers. 24 questions, about 10 minutes, instant results with your profile and your highest-risk layer.
No account needed. Your data stays private.