Skip to main content
MedMitra integrates with several powerful external services to deliver its AI-powered medical case management capabilities. Each integration plays a crucial role in the system’s functionality.

​
Core Integrations

Supabase

Backend-as-a-Service providing database, authentication, and file storage

Groq

Ultra-fast LLM inference for medical insights and radiology analysis

LlamaParse

Intelligent PDF parsing for extracting structured data from lab reports

Gladia

Real-time speech-to-text for medical dictation and note-taking

​
Integration Architecture

The integrations work together to create a seamless workflow:

​
Setup Requirements

Before using MedMitra, you’ll need to set up accounts and obtain API keys for:
  1. Supabase Project - Database, auth, and storage
  2. Groq API Key - LLM inference
  3. LlamaParse API Key - Document parsing
  4. Gladia API Key - Speech-to-text (optional)
All API keys should be stored in environment variables. See individual integration pages for specific setup instructions.

​
Environment Variables

MedMitra uses environment variables to securely manage API keys and configuration:

​
Backend (.env)

# Supabase
SUPABASE_URL="https://your-project.supabase.co"
SUPABASE_SERVICE_ROLE_KEY="your-service-role-key"

# AI Services
GROQ_API_KEY="your-groq-api-key"
LLAMAPARSE_API_KEY="your-llamaparse-api-key"

# Optional: Vector Database
WEAVIATE_API_KEY="your-weaviate-api-key"
WEAVIATE_REST_URL="your-weaviate-url"

​
Frontend (.env.local)

# Supabase (Client-side)
NEXT_PUBLIC_SUPABASE_URL="https://your-project.supabase.co"
NEXT_PUBLIC_SUPABASE_ANON_KEY="your-anon-key"

# Backend API
NEXT_PUBLIC_FASTAPI_BACKEND_URL="http://localhost:8000"

# Speech-to-Text
NEXT_PUBLIC_GLADIA_API_KEY="your-gladia-api-key"
Never commit .env or .env.local files to version control. Use .env.example files to document required variables.

​
Data Flow

​
Case Creation & Document Processing

  1. User uploads documents → Stored in Supabase Storage
  2. PDF lab reports → Processed by LlamaParse
  3. Radiology images → Analyzed by Groq Vision (LLaVA model)
  4. All data → Stored in Supabase Database

​
AI Insights Generation

  1. Aggregated data → Sent to Groq LLM
  2. Medical analysis → Case summary, SOAP notes, diagnosis
  3. Results → Stored in Supabase Database
  4. Display → Retrieved and shown in frontend

​
Speech-to-Text Workflow

  1. User starts dictation → Connects to Gladia WebSocket
  2. Audio streaming → Real-time transcription
  3. Final transcript → Added to case notes

​
Cost Considerations

Each integration has different pricing models:
  • Supabase: Free tier available, pay for storage/bandwidth
  • Groq: Usage-based pricing (very competitive rates)
  • LlamaParse: Credit-based system
  • Gladia: Pay per audio processing minute
Start with free tiers during development. Monitor usage in each service’s dashboard before scaling to production.

​
Next Steps

Set up Supabase

Configure database, authentication, and file storage

Configure AI Services

Set up Groq and LlamaParse for AI processing

Build docs developers (and LLMs) love

Get started for freeTalk to us