Portable computer
A vintage Polaroid camera case holds the Raspberry Pi computer, screen, and folding keyboard. The photographs follow the case fitting, components, and assembled computer.
Developed with assistance from Claude.
Thesis · 2005
Four miniature rooms from my thesis, Glass Houses: Invisible Connections.
Bo lived in the housing project directly behind the school. His box has the least light of the four. Inside are a small couch with newspapers, a television on a brass cart, and a hat and pocket watch on a side table.
Glass Houses brings together people I knew at Ringling in four miniature rooms. Cable television became a way to explore the connections between their homes and the divisions of race and class.
Each shadow box recreates a real person’s room. I photographed each person to fit the scene and made the figures as Kodalith transparencies using an AB development process. The rooms combine purchased dollhouse furniture with paper, fabric, and found objects that I made or adapted to scale. The purchased miniatures were also part of the commentary on consumerism: versions of domestic life sold as objects.
The work also examined my own white privilege: being able to enter each home as a guest and move between very different social and economic circumstances.
My colleagues at Bob Knight Photo took this photograph. I worked with their graduation crews across Florida during college. At my own graduation, I sat backstage with people I knew.
Graduation · May 2005
Two weeks before this photograph, another driver pulled out in front of me as I was returning from the dollhouse supply store. I braked but hit their car. I finished Glass Houses while recovering from surgery and using a wheelchair, moving the chair backwards to position the 4×5 camera. The BFA show opened on schedule.
This experience began my disability advocacy. Years later, at thirty-nine, my autism and ADHD diagnoses gave me another way to understand the work. The questions of race, class, and access in Glass Houses were connected to patterns I had noticed long before I could explain them.
Rooms & lighting
My desk, shelving, and star ceiling, photographed with different colors of light.
Rooms & display
The gallery wall and living room.
Lighting & supports
The shooting space, backdrop supports, and lighting setup.
Lettering
I 3D-printed the lettering for the Barbie photo booth and made the sign’s cardboard backing.
The lettering being printed.
Set construction
I built, painted, and assembled this photo booth, including the printed sign. It came apart for transport and went back together on site.
I built and painted the frame for a photo booth setup.
Honey favors
The engraving setup and the assembled honey-jar favor.
Paper work
Favor boxes, tags, and stationery.
Desktop interface
Layouts from my desktop usage widget. These are screenshots of the project.
My first AI project on Linux
My first project made with AI began on Linux as a companion app for Manuskript. It watches my book word count, hatches a dragon as the words accumulate, and sends the dragon flying around the screen as I progress.
The blue egg tracks the manuscript word count and the distance to hatching. As the book grows, the dragon hatches and begins to fly around the screen.
Developed with assistance from Claude.
Game made with Claude
A recording from the game project.
Audio build in progress
A Barbie jewelry box being rebuilt with speakers and a circular display. These photographs show the case, components, and current assembly.
Camera prototype
The camera module, measurements, electronics, and enclosure fitting.
Early project materials
The compact and circular display board for this project.
Handmade miniatures
Miniature vases I made, displayed alongside pieces from my collection.
Collected miniatures
Pieces from my miniature collection.
Client tool
A tool that walks a website the way a visitor would, at desktop and phone widths, runs accessibility and search checks, and asks a vision model to judge each page against the brand. The result is a report a business owner can read and act on.
Plain language, grouped by what to fix first. Every finding names the page and what to change.
Every page is loaded in a real browser twice, once at desktop width and once at phone width. Automated rules check structure, labels and contrast; a script measures weight, images, metadata, keyboard focus and tap targets, and reads what the site tells search engines and AI crawlers; then a vision model judges screenshots against a written brand profile, with its model name and rubric version recorded on every finding. The report groups what it found into fix first, fix next and small things.
To have your site reviewed, use the accessibility assessment lane on the contact page: it asks for the address and what worries you most about it. The review runs from the studio and comes back as a single web page you can open anywhere.
Built with Claude Code. Part of a five-project AI portfolio.
