EchoDoc

💡 Inspiration

In today's healthcare landscape, doctors are increasingly overwhelmed with patient communications and appointment scheduling, leading to burnout and reduced quality of care. Studies show that primary care physicians spend over 50% of their workday on administrative tasks rather than direct patient care. The healthcare industry loses approximately $150 billion annually due to inefficient communication and scheduling systems.

🔍 What it does

EchoDoc revolutionizes patient care management by creating AI-powered voice clones of doctors that can handle routine patient communications and appointment scheduling autonomously. This system allows doctors to maintain personal connections with patients while reducing administrative burden.

• AI agents with cloned doctor voices handle patient calls and appointment scheduling
• Personalized patient profiles ensure contextually aware responses based on medical history
• Dynamic calendar integration for real-time appointment management
• Secure handling of sensitive medical information and patient records

🖥️ How we built it

Frontend: Next.js 15 with App Router, TailwindCSS, Shadcn UI

Backend: Express.js

Authentication: Clerk

Database: Supabase

AI Integration: ElevenLabs for voice cloning, Gemini AI for text generation and conversation.

Communication: Twilio for call handling

🤔 Challenges we ran into

One of the main challenges was implementing real-time voice cloning that maintains the doctor's natural speaking patterns while ensuring the AI responses remain medically appropriate and safe. We also had to carefully handle the complexity of dynamic calendar updates across multiple concurrent AI agent calls, ensuring no double-booking or scheduling conflicts occurred.

🎉 Accomplishments that we're proud of

We successfully created a system that seamlessly integrates voice cloning technology with AI-driven patient interaction, maintaining the personal touch of doctor-patient relationships while automating routine tasks. The system's ability to parse and utilize patient records for contextual responses while maintaining strict medical privacy standards is particularly noteworthy.

📕 What we learned

This project taught us how to effectively combine multiple cutting-edge technologies - from voice cloning to real-time AI text generation - while maintaining HIPAA compliance and medical ethics standards. We gained valuable experience in creating systems that handle sensitive medical data while providing practical utility for healthcare providers.

⭐ What's next for EchoDoc

Future developments include:

• Expanding the system to handle more complex medical conversations
• Adding support for multiple languages
• Implementing more sophisticated appointment optimization algorithms
• Developing emergency situation detection and escalation protocols
• Creating a mobile app for doctors to monitor and manage their AI agents
• Integrating with existing Electronic Health Record (EHR) systems
• Continuing to build the calendar integration system.

How to run it: cd backend/express-server then run npm install then npm run dev

then another terminal, cd frontend then run npm install then npm run dev

then in another terminal you'll need to run ngrok or unicorn to expose the backend server so twilio can actually call it.

then fill in the environment variables in an .env file in backend folder.