My AI Approach

I see AI as a collaborator that, when designed responsibly, helps people move faster while staying in control.

In my own workflow, I use AI to uncover patterns in feedback, generate design variations, and synthesize complex information. I use it responsibly as an expert-in-the-loop and guide every step rather than handing over the wheel.

How I Use AI in My Daily Workflow

UI Variation and Stakeholder Alignment

I use Gemini Canvas to rapidly generate UI layout directions from a brief or dataset. It's useful for exploring multiple visual approaches quickly before committing to a direction in Figma.

AI-Augmented Diagrams

I chain NotebookLM and Miro together to build flowcharts faster without sacrificing accuracy. I prompt NotebookLM with uploaded project context to generate structured diagram content, then pipe that output into Miro's AI Diagram feature to produce the visual. I review, correct, reorient, and stylize the result manually to match my existing canvas.

Document Intelligence

I use NotebookLM as a personal RAG system. I upload resumes, course materials, and portfolio assets so I can query my own archive without reading through every document. It's grounded in my data, so it doesn't hallucinate.

Content and Communication

I use Claude and Gemini to draft portfolio copy, presentation narratives, and emails. I review the output critically and accept, rewrite, or reject it. AI accelerates the first draft and my judgment drives the final version.

Strategic Thinking Partner

I also use Claude with Skills and Gemini with Gems to validate ideas and stress-test assumptions before committing to a direction. It's not about outsourcing judgment. It's about sharpening it faster than I could working alone.

Research and Synthesis

I often use Perplexity Pro as my primary research tool instead of Google. I find Perplexity synthesizes results across multiple LLMs and surfaces cleaner, more actionable answers faster than traditional search.

I've also applied these principles to a full design project.

See the AI Photo Culler case study

Designing a Concept AI Agent to Cut Photo Culling Time by 50% While Keeping Photographers in Control

Photographers waste hours manually culling shoots. Through a structured AI/UX design program, I developed a concept prototype for Adobe Bridge that learns a photographer's style, auto-culls with explainable reasoning, and keeps the human in control.

  
  AI/ML
  Desktop (Native)
  B2B2C
  Conversational AI
  Multimodal AI

Role
Concept Product AI Designer

Target Culling
50% Time Saved

Established
5 Design Principles

Artifact Generation
13+ AI Tools Used

For audio voiceover walk through, unmute sound in the video.

VIEW CASE STUDY

AI-Generated Design Directions

Using a prompting framework learned from Jordan Wilson of the “Everyday AI” podcast, I generated three distinct dashboard directions in Gemini Canvas from a real financial dataset.

Each takes a different visual approach to the same data, reflecting the same method I'd use in a production environment to quickly surface options for stakeholder feedback before iterating toward a final solution.

Dashboard 1
Obsidian Ledger (dark, data-dense)

Dark, analytical aesthetic with anomaly detection and cash flow velocity visualization. Built for power users who want maximum data density.

Try Dashboard 1

Dashboard 2
Fiscal Architect (structural, minimal)

Light, structural layout using an efficiency ratio and stability matrix. Borrows from architectural drawing conventions to communicate financial health.

Try Dashboard 2

Dashboard 3
Financial Overview (clean, accessible)

Clear, approachable layout prioritizing scannability and plain-language labels. Designed for users who need quick answers without financial expertise.

Try Dashboard 3

Data anonymized and randomized before sharing. The original source was a real household financial spreadsheet tracking expenses, income, and net worth across multiple years.

Certifications

Master UX Design for AI Applications

Human Centered AI Institute · Aug 2025

Agentic AI and AI Agents Specialization

Prompt Engineering for ChatGPT
Agentic AI and AI Agents for Leaders
OpenAI GPTs

Vanderbilt University via Coursera · Jun 2025

AI Tool Stack

AI Tools: Used & Explored
Tool	How I Use It
Claude Pro Daily Use	Long-form writing, portfolio content, complex reasoning, and generating HTML/CSS code blocks for Squarespace portfolio visuals.
Gemini Pro Daily Use	Document intelligence, Google Drive integration, UI variation generation via Gemini Canvas, and generating HTML/CSS for portfolio components.
Perplexity Pro Daily Use	Primary research tool that synthesizes across multiple LLMs for cleaner, faster answers than traditional search.
NotebookLM Regular Use	Generates infographics and mind maps for portfolio visuals. I also use it to produce structured flowchart content that I pipe into Miro for visual diagram generation. As a personal RAG system, I upload my own documents and query them without hallucination risk.
Miro AI Regular Use	Used with the AI Diagram feature to convert structured NotebookLM output into visual flowcharts. I review, correct, reorient, and stylize the result manually to match my existing canvas.
ChatGPT Regular Use	Brainstorming, ideation, quick drafts, and generating HTML/CSS code for portfolio visuals and card layouts.
Voiceflow Explored	Explored for conversational AI prototyping, used to simulate chatbot flows in the AI Photo Culler project.
v0 Explored	Explored for UI prototyping, useful for research simulations but limited for pixel-accurate work.
Lovable Explored	Explored for UI generation, but couldn't replicate complex existing UIs faithfully.
Magic Patterns Explored	Explored for component-level UI generation. Found functional limitations for complex interfaces.
Figma Make Explored	Explored as AI-assisted prototyping, but pivoted back to traditional Figma for visual accuracy.

Design Tools

Figma

Axure

Miro

Sketch

Loom

Photoshop

Adobe Bridge

Arcade

Squarespace 7.1

< Disney Experience

Previous Work >