Big Red Macro

Big Red Macro is an intelligent, AI-driven dining planner built exclusively for Cornell University students. It seamlessly connects real-time data from the Cornell Dining API with your Google Calendar to generate hyper-personalized daily meal itineraries that perfectly balance your academic schedule, dietary restrictions, and macro goals.

Some Features

• Real-Time Cornell Dining Aggregation: Automatically scrapes the Cornell Open Data API to build daily local menus and cache hall operating hours.
• Smart Dining Dashboard: Explore all open and closed campus dining halls. View their active menus, check what dining plans they accept (Swipes vs. BRBs), and explicitly filter by campus area (North, West, Central).
• Personalized "Favorites" Network: Tap the heart icon on any menu item across any dining hall to add it to your profile's favorite_meals.
• Automated AI Meal Planner:
  • Sync securely with Google Calendar using an official OAuth 2.0 PKCE flow.
  • The backend leverages the Gemini 2.5 Flash LLM acting as a LangChain RAG pipeline.
  • It analyzes your free-time gaps and forcefully constraints itself to plan out your Breakfast, Lunch, and Dinner—strictly targeting your dietary restrictions (e.g. Vegan, Halal) and aggressively pushing menu options you marked as your favorites.
• Cornell-Native Aesthetic: A premium UI utilizing deep Cornell Red gradients and modern glassmorphism.

Stack

• Frontend: Vue 3, Vite, Tailwind CSS (Custom Cornell Red .cornell-red), Pinia, Vue Router.
• Backend: Python, Django REST Framework, Django-MongoEngine.
• Database: MongoDB (Caching menus, user profiles, AI configurations).
• AI Infrastructure: Google GenAI SDK (gemini-2.5-flash), heavily engineered LangChain Prompt Templates.
• External Integrations: Google Calendar Data API, Cornell Open Data Initiative REST API.

Getting Started

Prerequisites

• Docker & Docker Compose — the recommended way to run everything.
• A Google API Key with the Gemini API enabled (required for the Vision pipeline and AI meal planner).
• A Google OAuth 2.0 Web client ID for Google sign-in. Add http://localhost:5173 to the client's authorized JavaScript origins in Google Cloud.

1. Environment Variables

Copy the example env file and fill in your keys:

cp backend/.env.example backend/.env

Then open backend/.env and configure:

Variable	What it is	How to get it
SECRET_KEY	Django signing key	Generate one: python -c "from django.core.management.utils import get_random_secret_key; print(get_random_secret_key())"
GOOGLE_API_KEY	Powers the Vision pipeline (Gemini) + Google Calendar integration	Go to Google AI Studio → Create an API key. Make sure the Gemini API is enabled.
GOOGLE_CLIENT_ID	Verifies Google sign-in credentials on the backend and configures the frontend sign-in button	Google Cloud Console → APIs & Services → Credentials → Create OAuth client ID → Web application. Add http://localhost:5173 under Authorized JavaScript origins.
GOOGLE_CLIENT_SECRET	Exchanges Google Calendar OAuth codes for calendar tokens	Copy it from the same Google OAuth Web client.
GOOGLE_CALENDAR_REDIRECT_URI	Calendar OAuth callback URL sent to Google	Default: http://localhost:5173/calendar-callback. Add this exact URL under Authorized redirect URIs for the same OAuth client.
MAPBOX_ACCESS_TOKEN	Campus map and walking directions	Go to Mapbox → Create a token (the default public token works).
MONGODB_URI	MongoDB connection string	Defaults to mongodb://localhost:27017/bigredmacro. Docker handles this automatically.
REDIS_URL	Redis connection for Celery task queue	Defaults to redis://localhost:6379/0. Docker handles this automatically.
CORNELL_DINING_API_BASE	Cornell Dining API base URL	Default: https://admin-now.dining.cornell.edu/api/1.0
CORS_ALLOWED_ORIGINS	Allowed frontend origins	Default: http://localhost:5173

2. Running with Docker (Recommended)

Docker Compose spins up 4 containers: MongoDB, Redis, Django Backend, and Celery Worker.

# Set your Google API key so docker-compose can pass it to containers
export GOOGLE_API_KEY="your-key-here"
export GOOGLE_CLIENT_ID="your-oauth-client-id.apps.googleusercontent.com"
export GOOGLE_CLIENT_SECRET="your-oauth-client-secret"
export GOOGLE_CALENDAR_REDIRECT_URI="http://localhost:5173/calendar-callback"

# Build and start all services
docker compose up --build -d

Then run first-time setup:

# Run database migrations (Django auth tables)
docker compose exec backend python manage.py makemigrations
docker compose exec backend python manage.py migrate

# Pull today's dining menus from the Cornell API
docker compose exec backend python manage.py shell -c \
  "from api.services.dining_api import CornellDiningClient; c = CornellDiningClient(); c.ingest_all_menus()"

# Create a local testing user
docker compose exec backend python manage.py createsuperuser

3. Running Locally (No Docker)

If you prefer to run without Docker, you need to install and start these services yourself:

Service	Install (macOS)	Default Port
MongoDB 7	brew install mongodb-community then brew services start mongodb-community	27017
Redis 7	brew install redis then brew services start redis	6379

Then install Python dependencies and run:

cd backend
pip install -r requirements.txt
python manage.py makemigrations
python manage.py migrate
python manage.py runserver

To start the Celery worker (handles background tasks like the 4 AM favorite-food alerts):

cd backend
celery -A bigredmacro worker -l info -B

4. No MongoDB Schema Changes Needed

MongoEngine handles schema dynamically — all collections (including notifications for the favorite food alerts) are created automatically the first time they are written to. No manual migration or collection setup is required.

Accessing the Subsystems

Service	URL
Frontend Dashboard	http://localhost:5173
Backend Django API	http://localhost:8000/api/
MongoDB	mongodb://localhost:27017
Redis	redis://localhost:6379

Viewing Your Databases

MongoDB (menus, users, meal plans, notifications)

Use MongoDB Compass (GUI):

brew install --cask mongodb-compass

Open Compass, connect to mongodb://localhost:27017, and click on the bigredmacro database. You'll see collections like dining_halls, daily_menus, user_profiles, meal_plans, notifications, etc.

Or from the terminal:

mongosh mongodb://localhost:27017/bigredmacro

SQLite (Django auth only — users, tokens, sessions)

This is the small backend/auth.db file used only for Django's auth system. You can open it with DB Browser for SQLite, or from the terminal:

sqlite3 backend/auth.db ".tables"