VisioGram is a web application where users can generate images based on text prompts using a custom image generation model hosted on serverless GPUs (via Modal). The app features an intuitive user interface built with React and Tailwind CSS, and includes real-time interaction capabilities, such as viewing and managing generated images.

AI-powered genetic variant analysis platform using Stanford's Evo 2 model to predict mutation pathogenicity. Built with Next.js, FastAPI, and Serverless GPU acceleration for real-time genomic research and clinical decision support.