🔐 Sentinel – AI-Powered Crowdsourced Safety Heatmap Platform

Sentinel is a full-stack, microservice-based web application that provides real-time, crowd-sourced safety insights. By combining AI-driven sentiment analysis, Generative AI summaries, and dynamic geospatial visualization, Sentinel empowers users to make safer, smarter decisions about their environment.

🌟 A portfolio-ready capstone project integrating frontend, backend, AI/ML, and cloud-ready architecture with real-world impact.

---

🎥 Live Demo

Once you record a short GIF of the live app (login → search/pin → submit review → heatmap update → AI summary → moderation) replace this placeholder with the actual GIF.

---

💡 Motivation & Impact

Sentinel was built to bridge the gap between technology and community safety, showcasing an end-to-end integration of frontend, backend, and AI/ML.

• AI/ML-Powered Insights:

  • Sentiment Analysis: Converts user reviews into a quantitative sentiment score (+1 = positive, -1 = negative).
  • Generative AI Summaries: Aggregates reviews to create concise, human-readable summaries of safety at any location.

• Community & Public Utility:

  • Provides real-time safety maps for citizens.
  • Includes moderation mechanisms to ensure safe, accurate community contributions.
  • Potential for government partnerships for urban safety and law enforcement surveillance.

• Portfolio Significance:

  • Demonstrates full-stack development (React + Node.js + Firebase) and AI integration.
  • Highlights real-world problem-solving and social impact potential.

Sentinel turns crowdsourced data into actionable safety intelligence.

---

🚀 Features

🌍 Map & Heatmap

• Users can search or pin locations on the map.
• Submit safety reviews with:
  • Star rating (out of 5)
  • Pre-filled prompts + additional comments
• Heatmap color-coded by sentiment score:
  • Green = positive
  • Red = negative

🤖 AI-Powered Microservices

• Sentiment-Inference Service: ML model scores reviews from +1 to -1.
• Summarization Service: Gen AI summarizes all reviews for the selected area.
• Moderation Service: Detects toxic or inappropriate content; flags reviews as pending_moderation until approved.

📝 Community & User Features

• Secure user login with Firebase Authentication.
• Personalized user profiles with review history and moderation status.
• Upvote/Downvote system to highlight helpful reviews.

---

## 🔗 System Architecture

Diagram Explanation:

1. User Flow: Login → Search/Pin location → Submit review
2. Backend Flow: Review goes through API → Database → AI microservices
3. Moderation: Toxic reviews are flagged and hidden until approved
4. AI Services:
   • Sentiment Analysis → Heatmap update (Red/Green)
   • Generative Summarization → Display AI-generated summary
5. End Result: Users see a color-coded heatmap + AI summary for safety decision-making

⚙️ Technologies

• Frontend: React, Vite, React Leaflet, react-router-dom
• Backend: Node.js, Express.js
• Database: Firebase Firestore (NoSQL)
• AI Microservices: Python, FastAPI
• Machine Learning: Hugging Face Transformers, PyTorch, Scikit-learn
• Authentication: Firebase Authentication
• Deployment (Planned): Docker, Docker Compose, GitHub Actions, Google Cloud Run

---

🏃 Getting Started

✅ Prerequisites

• Node.js (v20.11+ or v22.2+)
• Python (v3.8+)
• Firebase project with Firestore + Authentication enabled

⚡ Local Development Setup

1. Clone the repository

git clone https://github.com/YourUsername/sentinel-platform.git
cd sentinel-platform

2. Setup Backend

cd backend
npm install
# Add Firebase serviceAccountKey.json to /src/config
npm start

Backend runs at: http://localhost:5000

3. Setup AI Microservices

cd services/<service_name>
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
uvicorn main:app --reload --port <port_number>

• Sentiment Service → :8000
• Summarization Service → :8001
• Moderation Service → :8002

4. Setup Frontend

cd frontend
npm install
# Add Firebase client config to src/firebase.js
npm run dev

Frontend runs at: http://localhost:5173

---

🔮 Roadmap & Future Enhancements

• Proactive Safety Notifications: Real-time alerts for low-safety areas.
• Full Containerization: Dockerize all services for consistent deployments.
• CI/CD Pipeline: GitHub Actions + Google Cloud Run deployment.
• Community Integration: Partner with local authorities and civic organizations.

---

📂 Project Structure

sentinel-platform/
│── backend/                # Node.js Express backend
│── frontend/               # React + Vite frontend
│── services/               # AI microservices
│    ├── sentiment-inference/
│    ├── summarization/
│    └── moderation/
│── docs/                   # System diagrams & documentation
│── LICENSE
│── README.md

---

📜 License

MIT License – see LICENSE

---

🌟 Why Sentinel Stands Out

• ✅ Full-stack + AI/ML integration: End-to-end implementation.
• ✅ Real-world impact: Crowd-sourced safety insights, moderated for reliability.
• ✅ Portfolio-ready: Demonstrates modern technologies, microservices, and cloud-readiness.
• ✅ Social value: Can be extended to government/public safety collaboration.

Sentinel is a working, production-like platform turning data into actionable safety intelligence.