AI-Enhanced Practice
AI as a multiplier,
not a replacement.
I've integrated AI tools into every phase of my design practice, from architecture planning and research synthesis to documentation and developer handoff. Here's how it actually works in my workflow.
How I Think About AI in Design
AI compresses time, not judgment.
I use AI to move faster between raw data and shareable artifact. The design judgment, including what to prioritize, how to frame the problem, and what the user actually needs, still comes from me.
The output is only as good as the input.
Effective AI use in design requires knowing how to prompt, what to verify, and when to override the output. I treat AI like a very fast junior collaborator. Useful, but not autonomous.
Research synthesis is where AI earns its keep.
The most tedious and time-sensitive part of research is turning raw notes into patterns. AI tools like Dovetail dramatically accelerate this, letting me run more research cycles in the same timeframe.
Documentation and handoff are underrated use cases.
Designers often under-document. I use AI to accelerate the creation of design system specs, developer handoff packages, and technical documentation. It's critical but time-consuming to do well manually.
AI Tools in My Work: Real Examples
Used for: Architecture planning, design system documentation, developer handoff specs, copy and microcopy drafting.
Arrived at stakeholder reviews with complete platform integration mapping. Normally takes days of back-and-forth with engineering.
Used for: UX flow stress-testing for edge cases, failure-state scenario planning, failure-state UX rules documentation.
Generated the failure-state UX rules document covering insufficient wallet funds, address verification failures, and mid-payment abandonment.
Used for: Research synthesis: tagging, clustering, and surfacing patterns across interview transcripts and field notes.
Cut research synthesis turnaround by ~40%, enabling faster decision cycles across multiple business units simultaneously.
Used for: Behavioral analytics AI for identifying usage patterns across digital properties, prioritizing high-impact usability issues.
Identified which pages and flows had the highest-impact usability issues, directing research effort to highest ROI areas.
The point isn't the tools. It's what they enable.
In a 10-day product sprint, AI compression is the difference between shipping everything and cutting scope. In a long-running enterprise engagement, it's the difference between one research cycle per quarter and three. I use AI to do more research, not less, and to document that research well enough that it actually drives decisions.