-
-
-
RALPH!
-
Speech-to-Text Diagnostic Screen
-
Game & Waiting Room Screen
-
Staff Interface
Inspiration
Early warning signs are often the hardest to catch in young patients because they struggle to articulate what they are feeling, leaving critical risks hidden in busy waiting rooms.
What it does
RALPH is an interactive, AI-driven pediatric triage and continuous monitoring companion that bridges the communication gap between young children and care teams and turns unproductive waiting room time into life-saving clinical oversight.
How we built it
Frontend: React, TypeScript, Vite, Tailwind CSS
Backend: Node.js, Fastify, TypeScript
AI and voice: Google Gemini; browser speech APIs as a fallback
Database: PostgreSQL with Tiger Data/TimescaleDB
Other features: Socket.IO for live updates, i18next for multilingual translations, Framer Motion for animation
Hosting plan: Vercel with GoDaddy Domain
Challenges we ran into
- Figuring out the internal queuing system
Accomplishments that we're proud of
- Interactive, child-centered kiosk companion
- Speech-to-text and text-to-speech interaction
- End-to-end interface including both the patient and clinical side
- Web app and games deploy according to assessment
What we learned
- How to develop quickly!
- Setting clear boundaries in healthcare software to achieve safe clinical balance
Built With
- gemini
- node.js
- posgresql
- react
- typescript
Log in or sign up for Devpost to join the conversation.