AirWise

Your personal air-quality consultant— combining location data, health profile, and LLM to give you actionable recommendations for safe daily activities.

Comment

Inspiration

Air pollution is one of the leading environmental risk factors for lung cancer and other respiratory diseases, yet most air quality data is impersonal and hard to act on. We were inspired by the idea of giving individuals a personal health consultant for the air they breathe, turning raw numbers into meaningful, actionable guidance that helps them live and move safely every day.

What it does

AirWise is a mobile-first application that:

  • Shows real-time air quality risk cards personalized to each user’s health profile.
  • Lets users save key locations (home, school, workplace) for quick checks.
  • Provides AI-powered “Run Coach” recommendations for when, where, and how long to exercise safely outdoors.

How we built it

We built AirWise using Expo (React Native) for the frontend and integrated Google OAuth via Supabase for authentication and user management. Our Supabase PostgreSQL database stores user health profiles and saved places. For environmental data, we used OpenWeatherMapand AirNow for AQI and weather, and Google Maps API for geolocation and place management. To generate personalized activity recommendations, we connected health and air quality inputs to Gemini, delivering natural, contextual advice. Notifications are powered by Expo Push to keep users updated in real time.

Challenges we ran into

  • Authentication flow: Configuring Google OAuth and integrating Supabase securely took longer than expected.
  • Environment setup: Getting Expo, Node, and dependencies aligned across team members’ machines cost us valuable time.
  • React Native learning curve: Some teammates were new to mobile development, so getting comfortable with React Native and Expo took extra effort.
  • Time management: Balancing ambitious features with the tight hackathon timeline forced us to prioritize and cut scope.

Accomplishments that we're proud of

  • Built a fully functional mobile-first framework in the first day.
  • Created a smooth Google login → profile setup → personalized AQI → AI recommendations flow.
  • Integrated multiple services (Supabase, Google Maps, OpenWeatherMap, Gemini) into one cohesive product.
  • Designed a clean, user-friendly interface that makes complex data understandable at a glance.

What we learned

  • How to integrate authentication, database, and APIs into a unified app workflow.
  • How to balance technical ambition with hackathon time constraints, focusing on MVP.
  • The importance of prompt engineering and testing when building AI-driven features.
  • Effective collaboration dividing responsibilities and aligning on design decisions.

What’s next for AirWise

  • Expand support to include more pollutants (pollen, wildfire smoke, radon).
  • Develop more sophisticated AI coaching that adapts over time to user habits.
  • Add social features for sharing safe activity routes with friends.
  • Optimize backend caching and prediction for real-time scalability.
  • Deploy the app to application stores for public use.
Share this project:

Updates