Gio Jung

I’m a research engineer at YouDescribe Lab, an accessibility research group building video audio description to make video content accessible for blind and low-vision users. My work focuses on evaluating vision-language models as raters and describers, and building the multimodal ML pipelines that measure how well they actually perform.

I’m also a founding member and Lead Engineer of Mosaic Cognition, where we’re building an authoring tool that turns stories into structured comics to help autistic learners with reading comprehension. I lead the engineering side of the product, from system architecture to data and inference pipelines to the full-stack build.