Skip to main content

Architecture Overview

This document provides a comprehensive overview of Medispeak's architecture, explaining how different components work together to provide seamless voice-to-EMR functionality.

Component Overview​

1. Frontend Components​

Chrome Extension​

  • Technology: React + TypeScript
  • Key Features:
    • Voice capture and processing
    • EMR field detection
    • Real-time feedback
  • Integration Points:
    • Chrome Extension API
    • Backend API communication

Web Widget​

  • Technology: React + TypeScript
  • Features:
    • Embeddable interface
    • Customizable UI
    • Event-driven architecture
  • Integration:
    • Custom events system
    • PostMessage API

2. Backend Services​

Rails API Server​

  • Technology: Ruby on Rails
  • Responsibilities:
    • Request handling
    • Authentication
    • Business logic
    • Data persistence
  • Key Features:
    • RESTful API
    • WebSocket support
    • Background job processing
    • Caching layer

AI Processing Service​

  • Components:
    • Speech-to-text engine
    • Medical terminology processor
    • Context analyzer
    • Field mapper
  • Technologies:
    • TensorFlow/PyTorch models
    • NLP processing pipeline
    • Custom medical models

3. Data Storage​

PostgreSQL Database​

  • Schema Overview: 
    -- Core tables
    users
    organizations
    transcriptions
    emr_mappings

    -- Reference tables
    medical_terms
    field_templates
    specialties
  • Key Features:
    • Audit logging
    • Data encryption
    • Performance optimization

Redis Cache​

  • Usage:
    • Session management
    • Rate limiting
    • Real-time data
    • Job queues
  • Key Features:
    • High availability
    • Data persistence
    • Pub/sub capabilities

Data Flow​

1. Voice Input Processing​

sequenceDiagram
    participant U as User
    participant F as Frontend
    participant B as Backend
    participant AI as AI Service
    
    U->>F: Speak into mic
    F->>F: Buffer audio
    F->>B: Stream audio
    B->>AI: Process speech
    AI->>B: Return text
    B->>F: Send structured data
    F->>U: Show results

2. EMR Integration​

sequenceDiagram
    participant F as Frontend
    participant B as Backend
    participant EMR as EMR System
    
    F->>B: Send structured data
    B->>B: Validate data
    B->>EMR: Map to EMR fields
    EMR->>B: Confirm update
    B->>F: Success response

Security Architecture​

1. Authentication​

  • JWT-based authentication
  • OAuth2 integration
  • Multi-factor authentication
  • Session management

2. Data Protection​

  • End-to-end encryption
  • At-rest encryption
  • Audit logging

3. Network Security​

  • TLS/SSL encryption
  • API rate limiting
  • CORS policies
  • WAF integration

Scalability​

1. Horizontal Scaling​

  • Load balancing
  • Service replication
  • Database sharding
  • Cache distribution

2. Performance Optimization​

  • CDN integration
  • Query optimization
  • Caching strategies
  • Background processing

Monitoring and Logging​

1. System Monitoring​

  • Performance metrics
  • Error tracking
  • Usage statistics
  • Health checks

2. Logging​

  • Structured logging
  • Log aggregation
  • Audit trails
  • Analytics

Development Workflow​

1. Code Organization​

frontend/
├── src/
│   ├── components/
│   ├── services/
│   ├── hooks/
│   └── utils/
backend/
├── app/
│   ├── controllers/
│   ├── models/
│   └── services/
└── lib/
    └── ai/

2. Deployment Pipeline​

  • Continuous Integration
  • Automated testing
  • Staging environments
  • Production deployment

Future Architecture Considerations​

1. Scalability Improvements​

  • Microservices architecture
  • Serverless components
  • Edge computing
  • Global distribution

2. AI Enhancements​

  • Custom model training
  • Real-time learning
  • Specialty-specific models
  • Multi-language support

Contributing​

For information on contributing to Medispeak's architecture:

  1. Review our Contributing Guidelines
  2. Check Future Projects
  3. Join architecture discussions on GitHub