POCUS Learning & Scan Log Mini-Platform

A full-stack training platform for POCUS (Point-of-Care Ultrasound). Learners can enroll in courses and log scans. Faculty/Admins can manage courses and view analytics.

🧱 System Architecture Diagram

          +------------------------+
          |        USERS           |
          +-----------+------------+
                      |
                      v
     +----------------------------------------+
     |        AWS Amplify (Next.js Frontend)  |
     +--------------------+-------------------+
                          |
                          | HTTPS /api/*
                          v
          +------------------------------------------+
          |     CloudFront Reverse Proxy (HTTPS)     |
          +--------------------+---------------------+
                               |
                               v
       +---------------------------------------------+
       |  Elastic Beanstalk (Node.js Express API)    |
       +-------------------+-------------------------+
                           |
                           v
               +---------------------------+
               |     MongoDB Atlas         |
               +---------------------------+

🌐 Live Demo

Frontend (Amplify):
https://master.d1q4nf9qlja1a8.amplifyapp.com/login

👥 Test Accounts (Seeded Users)

Admin:
  Email: admin@test.com
  Password: password123

Faculty:
  Email: faculty@test.com
  Password: password123

Learner:
  Email: learner@test.com
  Password: password123

🚀 Tech Stack

Frontend

• Next.js 16 (App Router)
• React Query (TanStack)
• TailwindCSS
• Hosted on AWS Amplify
• API calls routed through CloudFront HTTPS Proxy

Backend

• Node.js + Express (TypeScript)
• MongoDB Atlas (Mongoose)
• JWT Authentication
• Zod validation
• Mongo aggregation pipelines (metrics)
• Deployed on AWS Elastic Beanstalk

Infrastructure

• Amplify → Static Next.js application hosting
• CloudFront → Reverse proxy for /api/*
• Elastic Beanstalk → Node backend hosting
• MongoDB Atlas → Managed database
• GitHub Actions → Backend CI/CD deployment

📁 Monorepo Structure

pocus-platform/
  backend/
    src/
      config/
      controllers/
      middleware/
      models/
      services/
      routes/
      validation/
      utils/
    tests/
    package.json

  frontend/
    app/
    components/
    api/
    lib/
    package.json

⚙️ Backend Setup

1️⃣ Install dependencies

cd backend
npm install

2️⃣ Environment Variables

Create backend/.env:

PORT=5000
MONGO_URI=mongodb://localhost:27017/pocus
JWT_SECRET=super-secret-key

Production (Elastic Beanstalk Environment Variables)

MONGO_URI=<Atlas connection string>
JWT_SECRET=<secure generated token>
CORS_ORIGIN=https://<AmplifyDomain>,https://<CloudFrontDomain>

3️⃣ Run backend in development mode

npm run dev

Backend runs at:

http://localhost:4000

🌱 Seed Database

npm run seed

Creates sample users and sample FAST course.

🧪 Run Backend Tests

npm test

🔌 Backend API Summary

Auth

POST /api/auth/register
POST /api/auth/login

Courses

GET /api/courses
POST /api/courses (faculty/admin)
PATCH /api/courses/:id (faculty/admin)

Enrollment

POST /api/enrollments
GET /api/enrollments/me

Scan Logs

POST /api/scan-logs
GET /api/scan-logs/me
GET /api/scan-logs (faculty/admin)

Metrics

GET /api/metrics/summary

🎨 Frontend Setup

cd frontend
npm install
npm run dev

Frontend runs at:

http://localhost:3000

🔧 Frontend Environment Variables

Create frontend/.env.local:

NEXT_PUBLIC_API_URL=https://<cloudfront-domain>/api

📦 Frontend Production Build

npm run build

📘 Deployment Notes

Frontend (Amplify)

• Static Next.js deployment
• API routed through CloudFront

Backend (Elastic Beanstalk)

• CI/CD with GitHub Actions
• Environment variables stored securely

CloudFront

• /api/* → EB backend
• HTTPS enforced
• Caching disabled for API routes

MongoDB

• Atlas IP allowlist includes EB servers and local IP

📙 Assumptions & Trade-offs

Implemented:

• Clean controller/service separation
• Zod validation per endpoint
• JWT auth + RBAC
• Proper error handling middleware
• Mongodb Aggregation pipelines
• Unit tests for core business logic
• Automated backend deployment

Trade-offs

• Because of interview time constraints, the system uses simple role-based access, not a granular ACL mechanism.
• Next.js frontend uses static export + client-side hydration, not full SSR, to ensure compatibility with Amplify Hosting.
• The backend is a single-node Elastic Beanstalk instance (sufficient for demo; in production multi-AZ load balancing is required).
• MongoDB Atlas networking uses IP allowlist for EB instance, not VPC peering (faster setup).
• Minimal error-handling and form validation included; can be expanded in production

Future Enhancements

• Image upload to S3
• Audit logging
• CI for frontend
• Infrastructure as Code (CDK)

👨‍💻 Author

Romeo Enso
Full-Stack Engineer — POCUS Learning Platform