Inspiration
Urban property decisions in emerging markets like Sri Lanka often rely on incomplete information. Buyers and renters struggle to assess neighborhood safety, environmental risks, and community dynamics before making life-changing investments. We built UrbanPulse to bridge this gap — giving users the pulse of any neighborhood through real-time data and community intelligence.
What it does
UrbanPulse is a data-driven real estate platform that combines:
- Smart Property Search — Filter homes and commercial spaces by type, price, and location
- Environmental Intelligence — Real-time weather overlays, flood risk mapping, and air quality data
- Community Signals — Crowdsourced reports on infrastructure issues, safety concerns, and neighborhood updates
- AI-Powered Insights — Chat with our AI assistant about properties, neighborhoods, and market trends
- Interactive Map Explorer — Visualize properties with route planning and nearby amenities
- Real-time Messaging — Direct communication between buyers, renters, and property owners
How we built it
Frontend: React 18 + TypeScript + Vite for blazing-fast performance, styled with Tailwind CSS and shadcn/ui components. TanStack React Query handles data fetching and caching.
Backend: Express 5 + TypeScript with Mongoose ODM connecting to MongoDB. RESTful API architecture with JWT authentication.
Maps & Data: Leaflet for interactive mapping, OpenWeatherMap API for environmental layers, OpenRouteService for navigation.
AI Integration: Gemini API powers our property assistant for intelligent Q&A.
Infrastructure: Fully Dockerized with Docker Compose for seamless development and deployment.
Challenges we ran into
- Integrating multiple map tile layers (weather, traffic, environmental) without performance degradation
- Building a real-time community reporting system with location verification
- Designing an intuitive UX that presents complex environmental data simply
- Implementing secure JWT authentication with proper token refresh flows
Accomplishments that we're proud of
- Created a fully functional real estate platform in 24 hours
- Integrated 5+ external APIs seamlessly (weather, routing, AI, maps)
- Built a responsive design that works flawlessly on mobile and desktop
- Developed a community-driven reporting system with moderation
What we learned
- The power of TypeScript for full-stack type safety
- Efficient Docker orchestration for multi-service applications
- Balancing feature richness with user experience simplicity
- Real-time data visualization techniques with Leaflet
What's next for UrbanPulse
- Machine learning for property price predictions
- Historical environmental data analysis
- Mobile app with push notifications for community alerts
- Expansion to other South Asian markets
Built With
- docker
- express.js
- gemini-api
- jwt
- leaflet.js
- mongodb
- mongoose
- node.js
- openrouteservice
- openweathermap-api
- react
- shadcn/ui
- tailwind-css
- tanstack-query
- typescript
- vite
Log in or sign up for Devpost to join the conversation.