PictureMe

Project Description

AI-powered event photo platform that helps attendees instantly find the photos they appear in.

Project Overview

Built a full-stack event photo retrieval platform that transforms large event galleries into personalized photo collections. Event organizers can create events, upload photos, and share a QR code or private link with attendees. Users can sign in, complete a face profile with 3–5 selfies, and receive a personalized My Photos gallery powered by AWS Rekognition face matching.

My Role

Full-Stack Engineer (Team Project)

The Problem

After events, attendees often have to search through social media posts, shared links, group chats, and galleries of 1,000+ photos just to find the pictures they appear in.

Key Contributions

• ✦Built a full-stack web platform connecting event access, user authentication, photo uploads, and AI-powered face matching into one seamless workflow.
• ✦Engineered a FastAPI backend that coordinates Supabase authentication, Cloudinary photo storage, and AWS Rekognition face indexing/matching.
• ✦Implemented a reusable face-profile flow where users enroll once with 3–5 selfies and can receive matched photos across events they join.
• ✦Developed an event-based QR/link join system so attendees can access only the private events they were invited to.
• ✦Designed personalized My Photos and full All Photos gallery views, reducing manual browsing through large event galleries by an estimated 90%.
• ✦Integrated role-based access across creators, admins, and members to control who can upload photos, manage events, and view galleries.
• ✦Added privacy-focused features including optional face enrollment, private gallery access, and 30-day media expiration for cost-conscious photo storage.
• ✦Won 1st place and $2,000 at HornetHacks 2026 by delivering a working end-to-end AI event photo retrieval platform.