Lost and Found -- Marvin Ridge High School

A full-stack web application that helps Marvin Ridge High School manage lost and found items. Students report items they have lost or found, browse what others have submitted, and work with school administrators to reunite people with their belongings.

This project was built for the FBLA Website Coding and Development competitive event.

---

The Problem

Schools deal with dozens of lost items every week. Paper-based lost and found systems are slow, disorganized, and hard to search. Students often never recover what they have lost because there is no easy way to check what has been turned in.

Our Solution

This application gives students a clean, searchable interface to report and find items. It uses AI to moderate submissions automatically, correct search spelling mistakes, and generate descriptions from uploaded photos -- reducing the workload on school staff while making the system faster and more reliable for students.

---

Features

For students:

• Report a lost or found item with a title, description, category, location, date, and optional photo
• Browse all approved items with filters for type (lost/found) and category
• Search using natural language with AI-powered spell correction
• Claim a found item by submitting proof of ownership
• Send an inquiry to administrators about a specific item
• View notifications for admin replies and claim updates

For administrators:

• Dashboard showing pending items awaiting review
• Approve, reject, or delete submitted items
• Review and process item claims
• Respond to student inquiries
• View system-wide statistics

AI capabilities:

• Text moderation on all submissions (flags inappropriate content before it reaches an admin)
• Image moderation on uploaded photos
• Intelligent search that handles typos and alternate phrasings
• Automatic image descriptions generated from uploaded photos

---

How It Works

1. A student creates an account with a username and password. The system handles authentication through Firebase behind the scenes.

2. To report an item, the student fills out a form with details and optionally uploads a photo. The backend runs the text through an AI moderation check (Groq Llama 3.1) and, if a photo is included, checks the image as well (OpenAI GPT-4o-mini). If the content passes moderation, the item is saved with a "pending" status.

3. An administrator reviews pending items in the dashboard and either approves or rejects them. Approved items become visible in the public catalog.

4. Other students can browse the catalog, use AI-assisted search to find specific items, submit a claim with proof of ownership, or send an inquiry to an admin.

5. Administrators manage claims and respond to inquiries. Students receive notifications when there are updates.

---

Setup and Installation

Prerequisites

• Python 3.10 or later
• Node.js 18 or later
• A Firebase project with Realtime Database and Email/Password Authentication enabled
• API keys for OpenAI, and Cloudinary

Backend

cd backend
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Copy .env.example to .env and fill in your API keys and Firebase service account credentials.

Start the server:

python main.py

The API will be available at http://localhost:8000. FastAPI's built-in documentation is at http://localhost:8000/docs.

To create an admin account:

python create_admin.py <username> <password>

Frontend

cd frontend
npm install

Copy firebase.env.example to .env.local and fill in your Firebase project credentials. Set NEXT_PUBLIC_BACKEND_URL to point to your backend (defaults to http://localhost:8000).

Start the development server:

npm run dev

The application will be available at http://localhost:3000.

---

Copyright

This project was created for the FBLA Website Coding and Development event. All rights reserved.