MedPrompt AI

Inspiration

Managing chronic conditions like Thalassemia can be overwhelming—especially in areas with limited healthcare access. We were inspired to build MedPrompt AI to offer patients a smart, accessible health assistant that supports symptom checks, multilingual communication, emergency management, and connection with care services—all in one platform.

What it does

MedPrompt AI is a web-based health assistant that allows users to: Check symptoms using natural language. Connect with nearby hospitals or pharmacies. Access multilingual support using Lingo. Trigger emergency responses (like calling local services or alerting a saved emergency contact). Manage personal health data securely using blockchain.

How we built it

React and Tailwind CSS for frontend development. Supabase for backend database and authentication. Gemini AI for symptom interpretation. Lingo API for multilingual translation. ElevenLabs for voice output. Geo APIs to locate nearby hospitals/pharmacies.

Challenges we ran into

Voice response behaved unexpectedly on mobile browsers. Token and API usage limits required us to carefully manage requests. Integrating emergency calling in-browser had cross-platform limitations.

Accomplishments that we're proud of

Built a fully working, multilingual, AI-powered health assistant on a tight budget. Integrated real-world healthcare functionality with map-based and voice features. Deployed and tested across multiple platforms.

What we learned

Importance of designing around user constraints (e.g., browser support, mobile limits). How to optimize API token use and still provide full functionality. Managing health-related data securely with blockchain. UX considerations for emergencies and health accessibility.

What's next for MedPrompt AI

Add a dedicated blood donor matching module to help Thalassemia patients find and contact donors. Partner with organizations like Blood Warriors to support ongoing donation drives. Improve real-time emergency features with ride-booking integration. Launch a mobile-first version with better voice input/output performance.

Built With

• css
• elevenlabsapi
• geminiapi
• google-maps
• javascript
• lingoapi
• netlify
• react
• supabase
• tailwind
• tel
• webspeechapi