Project Overview

Healicure - Yejin Clinic

AI-powered EMR system for Korean medicine clinics

An AI prescription-recommendation EMR built for Korean medicine practitioners. As clinicians enter patient symptoms, the system suggests prescriptions in real time. AWS Amplify Gen1 multi-tenant auth and DynamoDB access-pattern optimization cut initial consultation (intake) time from 27 min to 12 min.

💡 Key Outcomes

• Faster intake: AI prescription recommendations reduced initial consultation time from 27 min → 12 min (55% reduction)
• TIPS selection contribution: Beta testing at 3 clinics with 30 intake logs contributed to TIPS program selection (May 2023)
• GraphQL optimization: Query optimization reduced network overhead by 60%
• DynamoDB key design: Access-pattern analysis and key schema optimization
• Multi-tenant auth: Amplify Gen1–based multi-tenant authentication and data isolation per clinic

🏗 Technical Architecture

1. Frontend & Auth

• ReactJS · TypeScript: EMR UI and real-time AI prescription recommendation interface
• AWS Amplify Gen1: Multi-tenant authentication and per-clinic data isolation
• Cognito: User authentication and authorization

2. Backend & Data

• Lambda · GraphQL: API and AI prescription logic; query optimization cut overhead by 60%
• DynamoDB: Access-pattern–driven key design for consistent low-latency reads
• CloudFormation: Infrastructure as code and deployment automation

🛠 Key Technologies

ReactJS, TypeScript, AWS Amplify Gen1, Lambda, DynamoDB, GraphQL, Cognito, CloudFormation

Developed with React, AWS Amplify Gen1, and AI-driven EMR for Korean medicine clinics.