|January 2026
CampusFind
ReactViteTailwind CSSFramer MotionPythonFlaskPostgreSQLOpenAIGemini AICloudinaryFirebaseVercel
Lost something on campus? CampusFind has you covered. CampusFind is a full-stack, AI-powered Lost & Found management platform built specifically for college campuses.
It replaces the chaos of WhatsApp group spam and physical notice boards with a centralized, intelligent system that makes recovering lost items fast, secure, and stress-free.
The core workflow is simple: Snap → Match → Verify → Recover. A student who finds an item snaps a photo and uploads it. The platform's AI vision engine instantly analyzes the image, auto-tags it with category, color, brand, and distinctive features, and proactively searches for matching "Lost" reports.
When a match is found, the built-in OTP verification system ensures the item is handed over to the rightful owner - no ambiguity, no fraud.
Key Features
- AI Vision Analysis : Upload a photo of any found item and the platform's vision agent powered by OpenAI GPT-4o-mini instantly parses it to extract structured metadata - brand, dominant color, item category, visual description, and distinctive features like scratches or stickers. No manual tagging required.
- Proactive AI Matching: The system doesn't just passively list items. It actively compares newly uploaded "Found" items against existing "Lost" reports using visual similarity powered by GPT-4o, returning ranked matches with a confidence score and reasoning.
- OTP Verification & Secure Handover: Claims aren't just accepted on faith. The system generates a unique 6-digit OTP for each verified claim. The item is only marked as "Returned" once the finder scans the claimer's code at the physical handover, closing the loop with an auditable trail.
- Poster Generator: Need a physical flyer? The platform generates a printable, beautifully designed "LOST" poster complete with item details and a custom QR code - all rendered entirely in the browser, no backend needed.
- Stats & Leaderboard: A community dashboard tracks items lost, found, and returned campus-wide. A leaderboard rewards students who return items most frequently, building a culture of trust and helpfulness.
- Scalable Infrastructure: Images are compressed and served via Cloudinary CDN for near-instant load times. The backend connects to a Neon Serverless PostgreSQL database that scales automatically from a handful of users to an entire university without infrastructure changes.
Why I built this
Every semester, college campuses lose hundreds of items - laptops, IDs, wallets, AirPods - and almost none get returned. Not because people don't care, but because the systems for reporting and recovering lost items are genuinely broken.
The existing "solution" at most colleges is to post in a WhatsApp group and hope for the best. These posts get buried in minutes. There's no structure, no verification, and no accountability. Anyone can claim anything. I wanted to build something that actually solved the problem end-to-end.
The hackathon format gave me a tight deadline and forced me to make hard prioritization calls, which made the engineering decisions sharper.
Beyond the practical problem, CampusFind was a deliberate challenge for me to integrate multiple AI APIs into a single coherent product - using GPT-4o for vision tasks, Gemini for natural language generation, and Firebase for real-time auth and to ship a polished, full-stack application with a Python backend and React frontend under time pressure.