Inspiration The idea for MediCall came from a real-world problem: many elderly people struggle to take their medications on time, which leads to deteriorating health, repeat hospital visits, and an increased burden on caregivers. Hiring full-time attendants is often unaffordable, especially in underserved regions.
We wanted to build an AI-powered system that hospitals and clinics could use to automate patient engagement — especially to remind patients about medications, appointments, and emergency care — without needing human agents to call them every time.
What It Does MediCall is a full-featured hospital call center management system that:
Registers and manages patient records and prescriptions
Books appointments with available doctors
Sends automated reminders via Bland.ai voice calls, SMS, and email
Uses custom cloned doctor voices (via ElevenLabs) for reminders to build patient trust
Tracks call outcomes, logs interactions, and displays real-time metrics for staff
How We Built It We used a modern full-stack approach:
Frontend: Next.js 13+ (App Router), React, TypeScript
Styling: Tailwind CSS and shadcn/ui components
Backend: MongoDB + Mongoose for structured data
Authentication: NextAuth.js with role-based access (Agent, Admin, Supervisor)
AI Voice Calling: Bland.ai API with ElevenLabs custom voices
Dev Tooling: Bolt for AI and API integrations
Deployment: Vercel for frontend, Railway for database hosting
Challenges We Faced AI Voice Integration: Mapping natural-sounding medical reminder scripts with Bland.ai and ElevenLabs voice API was tricky, especially when customizing voices for different doctors.
Webhook Handling: Capturing and updating call statuses in real time required building a robust webhook system.
Scalability: Ensuring real-time reminders didn’t overload the call queue was a challenge we tackled through staggered call scheduling.
User Experience: Designing a responsive, intuitive interface for both agents and administrators took multiple iterations.
What We Learned How to combine AI voice technologies with real-world healthcare applications
How to manage complex backend data relationships using MongoDB and Mongoose
The importance of real-time feedback loops (e.g., webhook-based call updates) in patient-facing apps
Building a production-grade full-stack app rapidly under hackathon constraints
Log in or sign up for Devpost to join the conversation.