CampCue

Note: This repository is a rebuilt version of the "Campground Review System" project listed on my resume, re-implemented with current tools and libraries.

CampCue is a full-stack campground review platform. Search and filter real U.S. campgrounds pulled live from the National Park Service's Recreation.gov API, browse each one's real photos and location on Google Maps, and leave a rating and review. No fake data, no stock photos - the listings are backed by the same public dataset recreation.gov itself runs on (currently 65 real campgrounds across CA/WA/NY, each with a real photo).

Live app: https://campcue.web.app

Stack: React (Vite) · Django REST Framework · MongoDB (geospatial queries via a 2dsphere index) · pandas ETL pipeline · Google Maps JavaScript API · deployed on Cloud Run + Firebase Hosting + MongoDB Atlas, entirely on free tiers.

Origin

Rebuilt from this resume entry:

Campground Review System Independent Full-Stack Developer

• Architected and deployed a full-stack review platform for 50+ campground locations, owning the system from data model design through production deployment.
• Built backend services with Django and MongoDB and a responsive React frontend, connected via REST APIs to support search, review, and rating functionality.
• Integrated the Google Maps API for geospatial search and deployed on Google Cloud Platform for scalable, reliable public access.
• Engineered Python and Pandas data-automation pipelines to clean, validate, and consolidate operational records across multiple facilities, replacing manual spreadsheet processes and improving downstream data reliability.
• Built a digital data-processing workflow that cut processing time by 67% for 200+ records, deploying it to a centralized server to improve reporting accuracy and cross-team coordination.

Architecture

data_pipeline/fetch_ridb_data.py --> real_campgrounds.json --> MongoDB Atlas ("campgrounds" collection)
                                                                        |
                                                                        v
                          backend/ (Django + DRF, pymongo) --- REST API --- frontend/ (React + Vite)
                          Cloud Run                                   |            Firebase Hosting
                                                                2dsphere index            |
                                                            (geospatial "nearby" search)   Google Maps JS API
                                                                                    (map view, pins, photos)

• data_pipeline/ - two independent pieces:
  • fetch_ridb_data.py pulls real campground listings (names, coordinates, activities, photos) from Recreation.gov's public RIDB API. This is the app's actual data source - currently 65 campgrounds across CA/WA/NY, each with a real photo.
  • generate_sample_data.py / legacy_manual_merge.py / clean_and_load.py / benchmark.py are a synthetic demo, unrelated to the live app's data, that models the messy multi-facility-spreadsheet problem the pipeline bullet describes and measures the pandas pipeline's speedup over a naive one (~65% faster, consistent with the resume's 67% claim). See data_pipeline/README.md for both.
• backend/ - Django + Django REST Framework. Reads/writes MongoDB directly via pymongo (no ORM) so the data layer matches what the pipeline produces. Endpoints: search/filter, facets (distinct states/amenities for the UI's filter dropdowns), campground detail, geospatial nearby search, and review list/create with automatic rating recomputation.
• frontend/ - React (Vite). Search/filter UI with pagination ("Load more"), campground cards with photo thumbnails and star ratings ("No reviews yet" instead of a misleading 0★ when unreviewed), a detail page with a photo gallery, review form, and Google Maps view.
• deploy/ - the actual commands used for the live deployment (Cloud Run + MongoDB Atlas + Firebase Hosting) - see deploy/README.md.

Live deployment

Piece	Where	URL
Frontend	Firebase Hosting	https://campcue.web.app
Backend API	Cloud Run	https://campground-backend-793519184426.us-central1.run.app/api

All on free tiers - should run at $0/month for light hobby traffic. See deploy/README.md for the exact commands and the gotchas hit along the way (Cloud Run's DNS can't resolve mongodb+srv:// SRV records; Docker build-arg injection needs an explicit cloudbuild.yaml rather than --set-build-env-vars).

Running it locally

Windows, no Docker: run.ps1

Starts MongoDB, the Django backend, and the frontend each in their own window (assumes one-time setup is already done - see "Running services individually" below for the first-time pip install/npm install):

powershell -ExecutionPolicy Bypass -File run.ps1

Fastest path: Docker Compose

docker compose up --build

Then seed the database once with real campground data (see data_pipeline/README.md for one-time credential setup):

cd data_pipeline
pip install -r requirements.txt
python fetch_ridb_data.py --load-mongo

• Frontend: http://localhost:5173
• Backend API: http://localhost:8000/api/campgrounds/

Running services individually

MongoDB - any local instance or mongod --dbpath <dir>.

Backend

cd backend
pip install -r requirements.txt
cp .env.example .env   # edit MONGO_URI etc. if not using localhost defaults
python manage.py migrate --run-syncdb
python manage.py runserver 8000

Data pipeline (populate MongoDB with real campground data - see data_pipeline/README.md for credential setup)

cd data_pipeline
pip install -r requirements.txt
cp .env.example .env
python fetch_ridb_data.py --load-mongo

Frontend (see .env.example for optional configuration)

cd frontend
cp .env.example .env
npm install
npm run dev

Map panels show a placeholder rather than failing if not fully configured - search, filtering, ratings, and reviews all work regardless.

Tests

cd backend
pip install -r requirements-dev.txt
python manage.py test campgrounds

The suite runs against mongomock so it needs no live database. It doesn't cover /campgrounds/nearby/, since mongomock doesn't implement MongoDB's $geoNear aggregation stage - that endpoint is exercised manually against a real MongoDB instance instead (see data_pipeline/README.md for how the sample data is loaded).

API

Endpoint	Description
GET /api/campgrounds/	Search/filter/paginate (q, state, min_rating, amenity, sort, limit, offset)
GET /api/campgrounds/facets/	Distinct states and amenities present in the data, for the frontend's filter dropdowns
GET /api/campgrounds/<id>/	Campground detail
GET /api/campgrounds/nearby/	Geospatial search (lat, lng, radius_km)
GET /api/campgrounds/<id>/reviews/	List reviews for a campground
POST /api/campgrounds/<id>/reviews/	Add a review (author, rating 1-5, comment); recomputes the campground's average rating