Skip to Main Content

cjrackley

Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

cjrackley — Demo video

My project is iVouch lite. It is a website that allows users to submit incidents in their area. The website requires an image and the user's location; this is so the image can be quickly scanned by the clip model to detect incident type and severity. In the lite form, this isn't optimized so there is a human checking aspect to it at well. A certified reviewer can look at posts made by users and confirm/deny whether or not it is a real post, getting 'vouches' for being real and removal if fake. This website was intended to include several models to correctly identify everything in the photo and matching it to its spot on the map. The original development plans and specs include android app implementation as well as rules and regulations for legal aspects including blurring faces, detecting sensitive material like CP/CSAM/DM and more. Due to time, budget and manpower it had to be cut down into its lite form.

Team