Work

Oracle Senior User Experience Designer, IC3 · Redwood City, CA

2023 to present

I own the behavior layer of AI products: the prompts, skills, contracts, and evals that decide what an agent says, how it formats it, and whether it stays in scope. Across all of it, my focus is standardization.

Featured

Shipped

Fusion Claw: Outcome Framework

The Outcome framework behind Fusion Agentic Applications: structurally consistent agent outcomes, in Redwood, across four product lines.

Public

Guideline: building high-quality AI responses

The standard for what a good AI response looks like, shipped with prompt blocks teams can enforce.

More work

Internal

Metrics and evaluations for conversational quality

Single-turn and multi-turn scoring on readability, completeness and word count, with test-case methodology.

Internal

AI Agent Studio: reusable skills

Plug-in skills for bias, voice and tone, and formatting, plus a visual evaluator that scores generated pages.

ProjectInternal

Voice and Mobile App

Starting an Outcome by voice delegation. I own the rules of voice, tone, and language, and extend internal best practices so components translate into voice.

Internal project. Happy to walk through it in an interview.

Internship · 2022

Unconscious bias in conversational AI

A user study with 12 participants on how people talk to Oracle’s digital assistant, and guidelines for a more inclusive model.

Much of this work is under NDA, so a lot of it stays high-level here. Happy to go deeper in a conversation.