Boston Daddy

main dashboard
weather and solar suggestions
important buildings
3D map

Inspiration

Boston faces critical infrastructure challenges that threaten the city's future. With 5 new electrical substations needed by 2035, 208 vulnerable stormwater outfalls, and traffic delays ranking 8th globally, we were inspired by the city's urgent need for smarter infrastructure management. Current systems operate in silos - energy grids, traffic management, and building systems don't communicate, leading to reactive rather than predictive problem-solving. We wanted to create a unified platform that could identify problems before they become disasters.

What it does

Boston Daddy is an AI-powered smart city platform that integrates energy, traffic, weather, and more to help cities worldwide make proactive, data-driven infrastructure decisions. The platform analyzes municipal energy consumption, traffic patterns, and building data across 15 buildings and 10 intersections to generate location-specific problem identifications with data-driven explanations. Our backend collects real-time data from municipal buildings, traffic intersections, and weather stations. The AI engine finds hidden correlations — for example, high traffic at Massachusetts Avenue intersections correlates with 23% higher energy usage in nearby buildings due to occupancy patterns. It provides priority-based problem ranking (high → medium → low) and an interactive dashboard with maps and charts for city officials. The current MVP demonstrates the foundation for a comprehensive smart city solution that will eventually include stormwater monitoring, predictive analytics, and citizen reporting - addressing Boston's critical infrastructure challenges from energy grid optimization to climate-resilient infrastructure planning.

How we built it

We built Boston Daddy using a modern full-stack architecture with Flask + Python backend serving RESTful APIs to a React + TypeScript frontend. The backend integrates Supabase (PostgreSQL) for data storage, Google Gemini AI for intelligent problem identification, and synthetic data generation for realistic municipal building and traffic data. The frontend uses Vite for fast builds, Tailwind CSS + Radix UI for beautiful components, Mapbox GL for interactive city visualization, and Zustand + TanStack Query for state management and data fetching. We implemented custom prompt engineering for the Gemini API to generate location-specific problem descriptions with data-driven explanations, creating a scalable platform that can easily expand to include stormwater monitoring, predictive analytics, and citizen reporting features.

Challenges we ran into

The biggest challenge was integrating AI-generated problem identifications with our existing data structures - we had to iterate through multiple prompt engineering approaches to get Gemini to generate consistent, location-specific problem descriptions while maintaining data accuracy. Setting up proper CORS configuration between frontend and backend for production deployment required careful Flask-CORS configuration, and creating realistic synthetic municipal data without access to actual city data demanded extensive research into Boston's infrastructure patterns. We also faced TypeScript build errors when transitioning from generic insights to location-specific problem identifications, requiring us to update frontend types and components to match the new AI-generated data structure.

Accomplishments that we're proud of

We successfully built a fully functional MVP that demonstrates AI-powered city problem identification with location-specific insights and data-driven explanations. We created a scalable architecture that seamlessly integrates multiple data sources (energy, traffic, weather) in a unified platform, developed an intuitive user interface that makes complex city data accessible to municipal officials, and achieved real-world deployment with proper environment configuration and production-ready code structure. Most importantly, we implemented a system that provides actionable problem identifications rather than generic recommendations, positioning Boston Daddy as a foundation for comprehensive smart city infrastructure management.

What we learned

We learned the importance of modular API design for scalable city data management and how to effectively integrate AI APIs for domain-specific problem identification. Creating realistic synthetic data for municipal infrastructure required extensive research into Boston's infrastructure patterns, and we discovered that balancing technical complexity with user experience in municipal software demands clear data visualization for non-technical city officials. We also learned that starting with a focused MVP before expanding to the full vision is crucial for managing complexity while demonstrating the core concept effectively.

What's next for Boston Daddy

The next phase involves adding real-time stormwater monitoring of Boston's 208 outfalls with AI-powered flood warnings, followed by implementing 48-hour forecasting for energy demand and traffic patterns. We'll then build a mobile app for citizen reporting of infrastructure issues and create a natural language interface for city officials to query the system. The long-term vision is to scale to the entire Massachusetts metro area, integrate with existing city systems (MBTA, Eversource, emergency services), and develop partnerships with other highly-populated cities for platform adoption, making Boston Daddy the comprehensive infrastructure management platform that every smart city needs.