Where the idea came from
During a spring heat wave, temperatures reached 85°F for three days while our dorm's air conditioning was still off. I felt the effects myself and saw how hard it was on the people around me.
Weather apps send the same heat alert to everyone, but people react to heat very differently. I wanted an app that learns each person's own sensitivity.
Who it's for
The main users are heat-sensitive people: those with conditions such as migraines or POTS, athletes and outdoor workers. It's also useful for anyone during a heat wave.
How it works
Users log the temperature, humidity, time outdoors, hydration, activity level, symptoms like fatigue, headache and dizziness, and sleep quality. The app calculates the heat index and a personal risk score, and shows trends over time.
The system has three parts. The brain finds patterns between conditions and symptoms. I based its rules on conversations with friends and classmates about how heat affects them. The skin is the dashboard. The coach gives advice and then asks follow-up questions to learn what actually helps.
The coach
Preventive: "You're approaching your typical fatigue threshold. Drink water and eat a light snack."
Reactive: "You reported dizziness and high heat. Move to a cooler environment immediately."
Adaptive: "You tend to underhydrate in the afternoon. Would you like to set a reminder?"
After each tip, it asks whether the symptoms got better, worse or stayed the same, and adjusts future advice.
What's next
Logging by hand is easy to forget and can be inaccurate. The next step is connecting wearables like an Apple Watch or Fitbit to track heart rate, skin temperature, sleep and activity automatically. Users can already export their logs to share with a doctor, and a future version could alert the doctor directly when it spots a high-risk pattern.
What I learned
AI was a great tool for building the app, since I don't have the coding skills to make one from scratch. It was much less useful for coming up with the idea. The ideas it suggested felt surface-level, and the real idea came from something I lived through.