Personal projects first, then earlier team and university work.
Guests scan one code and add photos, videos, and voice notes to an event’s gallery, with no app to install and no account to create. Photos are resized to WebP and stripped of EXIF and GPS data before they’re stored. Videos are transcoded and watermarked with ffmpeg on a separate background queue, so a long encode never holds up an upload someone is waiting on.
Documents move through an asynchronous pipeline for extraction, validation, facial verification, categorization, and routing before they reach a staff review queue. Azure Document Intelligence and OpenAI handle document understanding to cut down manual review. Queue state, validation results, and workflow tracking sit behind a Django and HTMX interface for staff.
Profiles, posts, follows, comments, shares, and likes work across independently run servers. I built authentication, follow and unfollow, the GitHub activity feed integration, and parts of the federated REST API, and wrote the Swagger documentation.
Connects Computing Science instructors with the students assigned as their TAs, and with students who want to TA in a future term.
Lets patients order their prescriptions and keep track of refills and order status without a trip to the pharmacy.
Upload a set of student photos named after each student, add the school logo, and download a finished PowerPoint deck with one slide per graduate.
Estimates the probability of diabetes from glucose level, blood pressure, body mass index, and age.
appengine.justinmonteza.comSource code for Diabetes Prediction
Tracks experiments and analyzes their results. Results are visible to every user by default, so people can follow and discuss each other’s experiments.
