AGENTIC
WEBSITE
BUILDERS
Three AI-native builders. Three different customer audiences. One team shipping all of them. The AI had to behave differently depending on who was using it while staying consistent enough to hold quality across all three.
Designers pushed PRs alongside engineering and PM from concept through release. Review cycles gave way to embedded guardrails and MCP-driven standards.
That operating model spread to other product teams across GoDaddy because the work got better, not because the process was mandated.
This site is a working example. Built entirely in code, directed by a single human in the loop. See how it was made.
- Took three builders from concept to front end release in two to three months each for separate customer audiences working as one team
- Designers work directly in Cursor and Claude Code, push PRs, and ship in production alongside engineering. I do this myself, not to model the behavior but because staying in the same friction keeps me calibrated.
- The boundary between design and front end development is gone. That changes what designers catch, what they fix, and how fast work moves. This operating model is now adopted across GoDaddy.
- Built a shared language system that taught the AI how our customers describe what they want when building a website. General language models do not know that vocabulary.
- Without it, every builder interprets the same customer input differently and drift compounds across products. What started as an interaction pattern became infrastructure across all three builders.
- $705K annual incremental GCR from UX-led Domains initiative, moving UX from advisory to accountable revenue partner
- Cannes B2B Grand Prix 2025 — led strategy and design of the Airo campaign site; changed the build approach to prove what the product could do with zero custom engineering
- Two generational shifts in one year, zero attrition — agentic design, then vibe designing; piloted with leads, expanded once the model proved out
- Operating model adopted by other product teams across GoDaddy — this team built it, the rest followed
- Led UX direction for Agent Name Service in partnership with OpenAI. ANS gives AI agents a verified identity on the internet. Working on it surfaced something that shapes how I think about agentic systems now: verification solves the identity problem but it does not solve the behavior problem. A verified agent can still act in ways that erode trust. That distinction changed how I approach system design.
- Normalized performance across a global UX org spanning North America and India — elevated India role to Director, established a Design Center of Excellence
- Mentored 10+ designers contributing to 4 hires and 3 promotions · Chair of Mentorship, AIGA Arizona
He expanded his scope meaningfully by shifting from execution-heavy ownership into experience strategy across agentic builder products, enabling his team to operate more autonomously while he focused on higher-leverage, cross-product problems. This included establishing new connections between UX, Research, and AI through initiatives like Common Language, and embedding UX more deeply into teams that had historically resisted design partnership.