NowCare

NowCare is an AI healthcare navigator that turns symptoms into safe next steps, then connects patients to nearby verified doctors, hospitals, ER status, scan slots, and appointments in real time.

Comment

Inspiration

When I first came to the United States as an international student, everything was new. The system, the environment, and especially healthcare. During my first semester, I broke my arm. What should have been simple became confusing. I did not know if I should go to the ER, urgent care, or wait to see a doctor. I did not understand how appointments worked or what to do next.

That experience stayed with me. I realized that for many people, especially students and immigrants, the hardest question is not "What is wrong with me?" but "Where should I go right now?"

That is what inspired NowCare.

What it does

NowCare is an AI-powered healthcare navigation platform that helps users understand the right level of care and connect with real healthcare options nearby.

Instead of trying to diagnose, NowCare guides users into one of six clear care paths:

  • Go to the ER now
  • See a doctor today
  • Get a scan
  • Try telehealth
  • Schedule a visit
  • Take care at home

After that, it helps users find:

  • Nearby verified doctors
  • Approved hospitals
  • Live ER status
  • Scan availability like MRI, CT, X-ray, and ultrasound
  • Appointment slots

Users can also upload medical documents and get simple summaries so they can understand them better before visiting a doctor.

How we built it

NowCare is a full-stack web app built using:

  • React, TypeScript, and Vite
  • Tailwind CSS
  • Firebase for authentication, database, and storage
  • Supabase for document storage
  • Google Gemini API for AI features
  • Google Maps API for location search
  • NPPES NPI Registry for doctor verification
  • CMS hospital data for hospital information

The app has different portals for:

  • Patients
  • Doctors
  • Hospitals
  • Admins

Firestore is used for real-time updates like bookings, availability, ER status, and patient activity.

Challenges we ran into

One big challenge was making sure the AI is helpful but safe. We did not want it to act like a doctor. So we limited it to only return clear care categories.

Other challenges were:

  • Designing one system that connects AI results with the user interface
  • Managing different user roles
  • Working with public data and real-time data together
  • Handling medical documents carefully
  • Keeping the app simple for users who may already feel stressed

Accomplishments that we're proud of

  • Built a working healthcare platform with multiple user roles
  • Used AI for symptom routing, document summaries, and assistance
  • Created real-time booking and availability system
  • Verified doctors using official NPI data
  • Added hospital approval system and ER status updates
  • Designed a safe AI flow without diagnosis

What we learned

We learned that in healthcare, AI should not try to sound smart. It should be clear, safe, and useful.

We also learned that real value comes from connecting everything into one flow:
symptoms → decisions → providers → availability → preparation

What's next for NowCare

  • Add insurance-based suggestions
  • Add ER wait time prediction
  • Let users share summaries with doctors
  • Add reminders and notifications
  • Improve accessibility and support more languages
  • Improve backend security for real use

Built With

Share this project:

Updates