Sebampitako/sarah_health_voice_agent

🏥 Sarah Healthcare Receptionist

Integrated AI-powered virtual receptionist for healthcare organizations using open-source models

Sarah is a professional virtual receptionist designed specifically for healthcare settings. She provides warm, efficient patient assistance using cutting-edge AI technology while maintaining the highest standards of professionalism and privacy.

🎯 FULLY INTEGRATED EXPERIENCE

Talk to Sarah directly in this Hugging Face Space - no external links required!

1. 1.Click "🚀 Start Sarah" above to activate the AI receptionist
2. 2.Click "Join Call" in the embedded video frame
3. 3.Allow microphone access when prompted
4. 4.Start talking! Sarah will greet you immediately

✨ Features

🤖 Professional Healthcare Reception

• —Warm, empathetic patient greetings
• —Appointment scheduling with complete information collection
• —Appointment cancellation and rescheduling
• —Office information (hours, location, parking, insurance)
• —Professional healthcare reception duties
• —Privacy-conscious and HIPAA-mindful

🚀 Open Source & Free

• —AI Model: Groq Llama 3.1 8B Instant (fast, free alternative to OpenAI)
• —Speech Recognition: Deepgram real-time transcription
• —Voice Synthesis: Cartesia neural text-to-speech
• —Video Platform: Daily.co WebRTC integration
• —Framework: Pipecat AI pipeline

🔧 Zero-Configuration Deployment

• —Pre-configured API keys for immediate use
• —Auto-starting AI assistant - just click Start Sarah
• —Embedded video calling - no external room links
• —Responsive design for desktop and mobile
• —Real-time status monitoring

💬 Example Conversations

Patient: "Hi, I'd like to schedule an appointment"

Sarah: "Hello! I'd be happy to help you schedule an appointment. Could you please provide me with your name and the type of appointment you're looking for?"

Patient: "My name is John Smith, I need a general checkup"

Sarah: "Perfect, John! For your general checkup, do you have a preferred doctor, and what date and time would work best for you?"

🏥 Sarah's Capabilities

📅 Appointment Management

• —Scheduling: Collects patient name, phone number, appointment type, preferred doctor, date, and time
• —Cancellation: Processes appointment cancellations with confirmation
• —Rescheduling: Helps patients find new appointment times
• —Confirmation: Provides clear appointment summaries

🏢 Office Information

• —Hours: Monday-Friday 8 AM to 6 PM, Saturday 9 AM to 2 PM
• —Location: Professional medical office details
• —Parking: Information about available parking
• —Insurance: General insurance acceptance information
• —Contact: Phone numbers and department directories

🤝 Professional Standards

• —Privacy Protection: No collection of sensitive medical information
• —HIPAA Awareness: Refers medical questions to licensed professionals
• —Warm Communication: Empathetic and professional tone
• —Efficient Service: Quick, helpful responses

🛠️ Technical Architecture

AI Pipeline

User Speech → Deepgram STT → Groq LLM → Cartesia TTS → User Audio

Core Components

• —Speech-to-Text: Deepgram WebSocket API for real-time transcription
• —Language Model: Groq's hosted Llama 3.1 8B with 150 token limit for speed
• —Text-to-Speech: Cartesia's neural voice synthesis (professional female voice)
• —Transport: Daily.co WebRTC for high-quality audio/video
• —Framework: Pipecat for AI pipeline orchestration

Performance Optimizations

• —Low Latency: Optimized audio buffers and sample rates
• —Fast Model: Llama 3.1 8B Instant for quick responses
• —Streaming: Real-time audio processing without delays
• —Smart Chunking: Efficient text-to-speech generation

🚀 Deploy Your Own

Hugging Face Spaces (Recommended)

1. 1.Fork this Space or create a new one
2. 2.Upload these files: app.py, requirements.txt, README.md, hf_spaces_config.py
3. 3.Set your API keys in hf_spaces_config.py
4. 4.Your Space will auto-deploy with Sarah ready to serve patients!

Local Development

git clone <this-repo>
cd sarah-healthcare-receptionist
pip install -r requirements.txt
python app.py

🔑 API Keys Required

The deployed version includes pre-configured API keys for immediate use. For your own deployment:

• —Groq API Key: Free tier available at groq.com
• —Cartesia API Key: Professional voice synthesis at cartesia.ai
• —Deepgram API Key: Speech recognition at deepgram.com
• —Daily.co API Key: Video conferencing at daily.co

🏥 Healthcare Use Cases

Medical Offices

• —Patient appointment scheduling
• —Office hour inquiries
• —General reception duties
• —Appointment confirmations

Clinics

• —Multi-doctor appointment coordination
• —Department information
• —Insurance verification support
• —Patient flow management

Telehealth Platforms

• —Virtual waiting room assistant
• —Pre-appointment information collection
• —Technical support triage
• —Patient onboarding

🔒 Privacy & Security

• —No Medical Advice: Sarah refers all medical questions to licensed professionals
• —Limited Data Collection: Only collects necessary appointment information
• —Temporary Sessions: No persistent storage of patient conversations
• —Professional Boundaries: Maintains appropriate healthcare reception scope

🌟 Why Open Source?

Sarah demonstrates that high-quality AI healthcare assistance doesn't require expensive proprietary models. By using open-source alternatives:

• —Cost Effective: Significantly lower operational costs than OpenAI-based solutions
• —Performance: Faster response times with optimized model selection
• —Transparency: Full visibility into AI model behavior and capabilities
• —Customizable: Easy to modify for specific healthcare organization needs

📞 Support

Built with ❤️ for healthcare professionals worldwide. This fully integrated experience makes AI-powered patient assistance accessible to any healthcare organization.

Ready to help your patients? Click "🚀 Start Sarah" above and begin the conversation! #Fixed