Hearth

Hearth is the Carfax for homes — an AI-powered platform that gives every property a verified health score, helping owners track value and buyers and renters shop with confidence.

Comment

About Hearth Property Management App

Hackathon Tracks

  • Pitch Competition
  • Demo Showcase

Inspiration

The inspiration for Hearth came from the growing need for intelligent property management solutions in today's real estate market. As property owners increasingly seek data-driven insights to maintain and optimize their investments, we recognized a gap in the market for a mobile-first application that combines AI-powered analysis with practical maintenance tracking.

The name "Hearth" reflects the core concept of creating a warm, central place where property owners can gather all their property information, much like how a hearth serves as the heart of a home. We were inspired by the complexity of property maintenance decisions and wanted to democratize access to professional-grade property analysis through AI technology.

The project was born from personal experiences with property management challenges - from unexpected maintenance issues to the difficulty of tracking multiple properties' health over time. We envisioned a solution that would not only track maintenance but also provide predictive insights to help property owners make informed decisions.

What it does

Hearth is a comprehensive property management mobile application that transforms how property owners interact with their real estate investments. The app provides:

🏠 Multi-Property Management: Users can add and manage multiple properties with detailed information including system ages, maintenance history, and property documents.

🤖 AI-Powered Analysis: Leveraging Cloudflare Workers AI, the app generates intelligent property health scores and recommendations based on system conditions, maintenance history, and predictive analytics.

📊 Dynamic Health Scoring: A sophisticated scoring algorithm that combines system ages, AI insights, and maintenance status to provide real-time property health assessments.

📅 Maintenance Timeline: A comprehensive tracking system for maintenance activities with status management (pending, in-progress, completed) and cost tracking.

💬 Interactive AI Chatbot: A floating chatbot that provides property-specific advice, local service recommendations, and can create pending tasks from conversations.

📄 Document Management: Upload and organize property documents, inspection reports, and maintenance records.

📱 Mobile-First Design: Built with React Native and Expo for seamless cross-platform performance on iOS and Android.

How we built it

Hearth was built using a modern, scalable tech stack designed for rapid development and future growth:

Frontend Architecture

  • React Native with Expo: Chosen for cross-platform development efficiency and excellent developer experience
  • Expo Router: File-based routing system for intuitive navigation structure
  • React Hooks & Context API: For state management and data flow
  • Custom Component Library: Reusable UI components with consistent design system

Backend & Data

  • Firebase Firestore: NoSQL database for real-time data synchronization and scalability
  • Firebase Authentication: Secure user management and session handling
  • Cloudflare Workers AI: Integration with @cf/meta/llama-3-8b-instruct model for property analysis

AI Integration

The AI analysis system was a core technical achievement:

// AI Analysis Service Architecture
const performAIAnalysis = async (propertyData: Property) => {
  const prompt = `Analyze this property and provide maintenance recommendations...`;

  const response = await fetch(CLOUDFLARE_API_URL, {
    method: 'POST',
    headers: {
      'Authorization': `Bearer ${CLOUDFLARE_API_TOKEN}`,
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      messages: [
        { role: 'system', content: 'You are a property maintenance expert...' },
        { role: 'user', content: prompt }
      ]
    })
  });

  return await response.json();
};

Health Score Algorithm

We developed a sophisticated scoring system that combines multiple factors:

// Health Score Calculation
const calculateHealthScore = (property: Property, aiInsights: AIInsight[]) => {
  const baseScore = 100 - (averageSystemAge * 2);
  const aiAdjustment = aiInsights.reduce((score, insight) => 
    score + (insight.impact * insight.weight), 0);
  const maintenanceAdjustment = resolvedItems * 2 - pendingItems * 3;

  return Math.max(0, Math.min(100, 
    baseScore + aiAdjustment + maintenanceAdjustment
  ));
};

Development Process

  1. Rapid Prototyping: Started with Expo CLI for quick iteration
  2. Component-First Development: Built reusable components before full features
  3. AI Integration: Implemented Cloudflare Workers AI early for testing
  4. Real-time Updates: Integrated Firestore listeners for live data synchronization
  5. User Testing: Continuous feedback loop with property owners

Challenges we ran into

1. AI Integration Complexity

Challenge: Integrating Cloudflare Workers AI with proper error handling and response parsing was more complex than anticipated.

Solution: We implemented a robust error handling system with fallback responses and detailed logging to ensure the app remains functional even when AI services are unavailable.

2. Real-time Data Synchronization

Challenge: Managing real-time updates across multiple components while maintaining performance proved challenging.

Solution: We implemented a custom hook system (useEnhancedHealthScores) with memoization to prevent unnecessary re-renders and ensure consistent data across the app.

3. Health Score Consistency

Challenge: Health scores were changing unpredictably when users navigated between screens, creating a poor user experience.

Solution: We developed a deterministic scoring algorithm with proper caching mechanisms and implemented useMemo to ensure scores only update when underlying data actually changes.

4. Firestore Concurrent Operations

Challenge: Multiple simultaneous Firestore operations were causing internal assertion errors.

Solution: We refactored all write operations to use Firestore batch writes for atomicity, with fallback to individual operations if batch operations fail.

5. Cross-platform UI Consistency

Challenge: Ensuring consistent UI behavior across iOS and Android while maintaining native performance.

Solution: We developed a comprehensive design system with platform-specific adjustments and extensive testing on both platforms.

Accomplishments that we're proud of

🎯 Technical Achievements

  • Seamless AI Integration: Successfully integrated Cloudflare Workers AI with robust error handling and response parsing
  • Real-time Architecture: Built a real-time data synchronization system that updates across all screens instantly
  • Intelligent Health Scoring: Developed a sophisticated algorithm that provides accurate, consistent property health assessments
  • Cross-platform Excellence: Delivered a native-feeling experience on both iOS and Android

🚀 Feature Innovations

  • Context-Aware Chatbot: Created an AI assistant that understands property context and provides location-specific service recommendations
  • Dynamic Maintenance Tracking: Built a comprehensive maintenance system with status workflows and visual indicators
  • Predictive Analytics: Implemented AI-powered property analysis that considers historical data and current conditions
  • Document Management: Seamless file upload and organization system for property documents

📱 User Experience

  • Intuitive Navigation: File-based routing system that makes the app feel natural and easy to navigate
  • Visual Feedback: Comprehensive loading states, animations, and status indicators that keep users informed
  • Accessibility: Built with accessibility in mind, ensuring the app is usable by a wide range of users

🔧 Code Quality

  • Type Safety: Full TypeScript implementation with comprehensive type definitions
  • Component Architecture: Reusable, well-documented components that promote maintainability
  • Error Handling: Robust error handling throughout the application with user-friendly error messages

What we learned

Technical Insights

  • AI Integration Best Practices: Learned the importance of proper prompt engineering and response validation when working with AI APIs
  • Real-time Data Management: Gained deep understanding of Firestore's real-time capabilities and the challenges of maintaining data consistency
  • React Native Performance: Discovered the importance of proper memoization and state management for optimal mobile performance
  • Cross-platform Development: Learned to balance native performance with code reusability in React Native

Product Development

  • User-Centric Design: The importance of continuous user feedback in shaping product features and user experience
  • Iterative Development: How rapid prototyping and iterative development can lead to better products faster
  • AI Product Integration: The challenges and opportunities of integrating AI into consumer applications

Team Collaboration

  • Documentation Importance: The critical role of comprehensive documentation in maintaining code quality and team collaboration
  • Testing Strategies: The value of both automated and manual testing in ensuring product quality
  • Performance Optimization: How to identify and resolve performance bottlenecks in mobile applications

What's next for Hearth

Immediate Roadmap (Next 3 months)

  • Enhanced AI Features: Implement more sophisticated AI analysis with historical trend analysis and predictive maintenance scheduling
  • Push Notifications: Add maintenance reminders and property health alerts
  • Offline Support: Implement local data caching and sync for offline functionality
  • Advanced Analytics: Detailed property performance metrics and reporting

Medium-term Goals (6-12 months)

  • IoT Integration: Connect with smart home devices for real-time property monitoring
  • Market Analysis: Integrate real estate APIs for market trend analysis and property valuation
  • Multi-language Support: Internationalization for global property owners
  • Web Application: Responsive web version for desktop users

Long-term Vision (1-2 years)

  • Predictive Analytics: Machine learning models for maintenance prediction and cost optimization
  • Property Comparison: Advanced comparison tools for portfolio analysis
  • Financial Integration: Expense tracking and ROI analysis features
  • Community Features: Property owner networks and shared maintenance resources

Technical Improvements

  • Performance Optimization: Advanced caching strategies and bundle size optimization
  • Security Enhancements: Two-factor authentication and advanced security features
  • API Expansion: Integration with additional property management and maintenance service APIs
  • Advanced Testing: Comprehensive test coverage with automated testing pipelines

Hearth represents just the beginning of our vision to revolutionize property management through technology. We're excited to continue building features that make property ownership more intelligent, efficient, and profitable for our users.

Built With

  • cloudfare-workers-ai
  • context-api
  • expo-application-services
  • expo-document-picker
  • expo-image-picker
  • expo-router
  • expo-vector-icons
  • expo.io
  • firebase
  • firebase-auth
  • firebase-firestore
  • react-hooks
  • react-native
  • react-native-stylesheet
Share this project:

Updates