A medical billing system with AI audit checks: React on CloudFront with a Lambda API, plus a self-hosted twin on Postgres. All data is synthetic. Demos: AWS and self-hosted. More in my personal projects.
Complete System Architecture
End-to-end view of the medical billing system showing user interaction, frontend, backend, and data persistence layers.
Medical Staff
End Users
→
CloudFront
CDN Distribution
→
React SPA
Frontend UI
→
AWS Lambda
FastAPI Backend
↔
PostgreSQL
Primary Database
↔
AI Services
Ollama + Bedrock
Frontend Architecture - React & AWS
React SPA deployed on AWS CloudFront with S3 static hosting and automated CI/CD pipeline.
AWS Infrastructure
CloudFront
d27z0qz3ducsem.cloudfront.net
→
S3 Bucket
medical-billing-frontend-v2
React Application
Home Page
Patients
Providers
Claims
Appointments
Audit
CI/CD Pipeline
GitHub
Source Code
→
Actions
Build & Deploy
→
Auto Deploy
S3 + CloudFront
Backend Architecture - FastAPI & Lambda
Python FastAPI application running on AWS Lambda with PostgreSQL database and AI integrations.
FastAPI Application
/api/patients
/api/providers
/api/claims
/api/payments
/api/audit
Database Layer
SQLLite (in memory) or PostgreSQL (formally)
Primary Database
patients
providers
claims
payments
audit_logs
services
Medical Billing Data Flow
Complete workflow from patient registration through claim processing to payment and audit.
Patient Registration
→
Provider Assignment
→
Appointment Scheduling
→
Service Delivery
→
Claim Creation
→
AI Audit
→
Insurance Processing
→
Payment Collection
→
Reporting
AI-Powered Claim Auditing
Advanced fraud detection and claim validation using Ollama LLM and AWS Bedrock AI services.
AI Services
AWS Bedrock
Production AI
Claude Models
GPT Integration
Scalable Processing
Enterprise Security
Ollama LLM
Local AI Server
Llama Models
Mistral Integration
Local Processing
Development Testing
Claim Submission
→
Pattern Analysis
→
Fraud Detection
→
Audit Report
Automated Deployment Pipeline
GitHub Actions-powered CI/CD for both frontend and backend deployments with automated testing and deployment.
Frontend Pipeline
GitHub Push
→
Vite Build
→
S3 Deploy
→
CloudFront Invalidation
Backend Pipeline
GitHub Push
→
Docker Build
→
ECR Push
→
Lambda Deploy
System Status & Deployment Info
Current deployment status, URLs, and configuration details for the medical billing system.
Frontend Status Active
CloudFront URL: d27z0qz3ducsem.cloudfront.net
S3 Bucket: medical-billing-frontend-v2-20250526
Build Tool: Vite
Framework: React 18 + React Router
Deployment: GitHub Actions
Cache Strategy: HTML: no-cache, Assets: 5min
Features: Navigation, Forms, AI Integration
Backend Status Active
Runtime: AWS Lambda (Python 3.9)
Framework: FastAPI + Mangum
Database: SQLLite (in memory) or PostgreSQL (formally)