At Kustomer, I rebuilt the design operating system around a new reality: designers, engineers, and AI systems working in the same loop.
Most design orgs were built for human-made work and software execution. At Kustomer, I redesigned the operating system for AI: designers, engineers, and models working together.
Days to prototype
Static design handoffs
Engineering interpreting specs
Same-day product exploration
Live AI interactions in design reviews
Working software instead of static artifacts
Direct testing with users earlier in the cycle
Impact Numbers
CONCEPT → WORKING PROTOTYPE
Time to testable direction
STATIC SPECS REPLACED
I identified a recurring gap in our AI workflow, built the first reusable UX checks myself, and scaled them into shared standards for
Design and Engineering.
Reusable AI skills helped Design and Engineering apply shared UX standards and make better product decisions.
# AUDIT UX: {FEATURE}
## What we encode
Undefined states
Dead ends
Abandonment risks
Recovery failures
Broken user journeys
## OUTPUT
Flow inventory + priority matrix. Critical → Low.
Turn design standards into reusable AI instructions.
## What we encode
Accessibility requirements
Content standards
Component usage
Brand voice
Error handling
Interaction patterns
## Output
Reviewable design specification + implementation guidance
A decision rule for where automation ends and human judgment begins.
## Decision tree
- Reversible / Low-Stakes → model decides, log it
- Reversible / High-Stakes → model proposed, human confirms
- Irreversible → human decides, model assists
## Default
Show your work (trust)
Review the AI behavior, not just flows and screens
## What we critique
- Outputs → are responses useful and correct?
- Recovery → what happens when the model fails?
- Confidence → can users trust the result?
- Boundaries → where does automation stop?
## Output
Design rules, prompt updates, evaluation criteria, and system improvements.
Work in live systems over static specs
Design prompts, outputs, rules, and recovery paths, not just flows.
Evaluate AI outputs against real use cases.
Co-create with engineering from the start.
Use real data and real user signals as early as possible.
Design the model’s behavior, not just
the interface.
Work in live loops: prototype, test, adjust.
Treat AI as part of the product system, not a side tool.
Codify what good looks like with skills, rules, patterns, and evals.
Co-create with engineering
from the start.
Use real data early.
Document decisions so teams can move independently.
I rewired my own practice first, then used that learning to help the team move from static design artifacts to AI-assisted prototypes, research synthesis, product critique, and shared evaluation systems.
Cursor
Claude Code/Design
Code-based prototyping
AI-assisted research synthesis and product analysis
// operating artifacts · tap to open
Codified common UX checks so teams could spend less time rechecking the basics and more time making the hard product calls.
ux-audit.md
Skill 01
Reusable AI skills transformed expert UX review into a repeatable system.
# AUDIT UX: {FEATURE}
Undefined states
Dead ends
Abandonment risks
Recovery failures
Broken user journeys
## OUTPUT
Flow inventory + priority matrix. Critical → Low.
ux-audit.md
Skill 01
Reusable AI skills transformed expert UX review into a repeatable system.
# AUDIT UX: {FEATURE}
Undefined states
Dead ends
Abandonment risks
Recovery failures
Broken user journeys
## OUTPUT
Flow inventory + priority matrix. Critical → Low.
output-eval.md
Skill 02
output-eval.md
Skill 02
human-in-the-loop
Skill 03
human-in-the-loop
Skill 03
critique.md
Skill 04
critique.md
Skill 04
Moved design team from static handoffs to AI-native co-creation: prototype with LLMs, test real outputs, iterate in hours.
Introduced Markdown-based skills to standardize AI usage across prompts, component rules, output evals, and human-in-the-loop guardrails.
Turned AI from individual experimentation into a shared operating system.
Moved design critique from subjective review to observable output quality. AI-generated work is evaluated against real use cases, defined edge cases, and documented product intent before shipping.
Eliminated traditional handoff by embedding design and engineering in the same loop: co-defining behavior, prototyping together, and reviewing outputs collaboratively.
Compressed idea-to-working-output timelines from days or weeks to hours, reduced interpretation gaps and rework, and tested design quality against real conditions earlier.
Creative thinking did not disappear. It moved from static files into system behavior: prompts, rules, evaluation criteria, and the user’s voice in the process.
Moved from static handoffs to AI-assisted co-creation: prototype, test, and iterate
in hours.
Created reusable skills, rules, and guardrails to standardize AI-assisted design work.
Evaluated outputs against real use cases, edge cases, and product intent.
Design and engineering worked in the same loop: prototyping, reviewing, and
refining together.
Reduced idea-to-output cycles from days to hours and decreased rework.
Design judgment moved from files into prompts, rules, evaluations, and user outcomes.
