Overview & Scope
Project Goals
Back in 2018, I joined a startup called AICA as their only developer. The goal was to use AR to help deaf and visually impaired people read public signs and books. I had never done AR or mobile development before, but it seemed like an interesting challenge.
For the MVP, we needed the app to recognize specific images and play sign language videos with transparent backgrounds over them. We also needed a way to upload new content without pushing app updates, plus a demo book we could use to show investors.
As we got going, we started adding more features - like a map showing accessible places, user reviews about accessibility, and a way to connect physical locations with their scannable content like restaurant menus.
Design Process
My first lessons
When I started, there was already a brand stylebook made by a graphic designer before me. They had created it for print materials, not really thinking about app development. So I had to figure out how to translate their color scheme, typography, and logo into something that would actually work on screens - which was an interesting puzzle since some of those colors had pretty poor contrast ratios.
I was just getting into UI design and Sketch at the time. The first version tried to stay faithful to the brand guidelines while still being usable:
After actually building and using it, I made a second version that leaned more into Material Design principles I'd discovered along the way. It bent the brand guidelines a bit more, but would've been much more usable. This version never got implemented though.

Technical Implementation
AR & Mobile Development
This project introduced me to a bunch of new tech. I had to learn Vuforia for the AR features, which somehow led me into OpenGL shader territory. The app itself was built with Android Studio and Java - both firsts for me - and Firebase handled all the backend stuff.
The most interesting challenge was getting transparent videos to play over the camera feed in AR, which required diving into GLSL shaders.
Working solo on this project pushed me to learn different development environments and experience Android development challenges. The project successfully reached MVP stage but ultimately didn't continue as the company ran out of money.