SmartHealthCare — Stunting Detection Platform

SmartHealthCare is a full-stack web application developed during an internship at Indihealth. The platform supports researchers in detecting and analyzing stunting cases using multiple AI-based classification methods, combined with geospatial data visualization.

GitHub/Gitlab

SmartHealthCare — Research Platform for Stunting Detection

SmartHealthCare is a web-based research platform developed during an internship at Indihealth. The system is designed to assist researchers and health professionals in detecting, analyzing, and visualizing stunting cases using multiple AI classification approaches and geospatial mapping across Indonesia.

The project combines a modern full-stack web application with an AI-powered backend, built collaboratively with a dedicated machine learning engineer.

Project scope

  • Role: Full-Stack Web Developer
  • Domain: Public Health, Medical Research, Data Analytics
  • Collaboration: AI/ML backend developed by a separate team member
  • Target users: Researchers, analysts, and healthcare professionals

Highlights

  • Multi-method AI-based stunting classification
  • Interactive geospatial visualization of stunting distribution
  • Clean separation between web frontend and AI inference services
  • Designed for research workflows and data exploration

Core features

  • Authentication & Authorization
    • Secure login system for researchers and administrators
  • Stunting Detection Module
    • Upload and process health data
    • AI classification results served via REST API
  • Analytics Dashboard
    • Summary statistics and classification outcomes
    • Dataset-level insights for researchers
  • Geospatial Mapping
    • Interactive maps using MapTiler
    • Visualization of stunting distribution across Indonesian regions
  • Data Management
    • CRUD operations for research datasets
    • Structured storage for patient and regional data

Tech stack

Frontend

  • React.js
  • Material Tailwind (UI components and layout)
  • REST API consumption for AI inference and data services

Backend & AI Services

  • Python
  • FastAPI (AI service layer)
  • Uvicorn (ASGI server)
  • AI / ML libraries for stunting classification (model experimentation and inference)

Database

  • MySQL
  • SQLAlchemy ORM

Mapping & Visualization

  • MapTiler for geospatial data visualization

System architecture overview

  1. Web Frontend (React)
    • User authentication
    • Dashboard, analytics, and map visualization
  2. Backend Services
    • REST endpoints for data management
    • AI inference endpoints exposed via FastAPI
  3. AI Classification Layer
    • Multiple classifier models for stunting detection
    • Model results returned as structured JSON
  4. Database Layer
    • MySQL with SQLAlchemy for relational data modeling

This modular architecture allows independent scaling and experimentation of AI models without affecting the frontend.

Technical challenges

  • Integrating AI inference services with a production-grade web frontend
  • Handling large research datasets efficiently
  • Designing map-based visualizations that remain readable at national scale
  • Maintaining clean API contracts between web and AI services

Impact & lessons learned

  • Close collaboration between full-stack and AI roles is critical in research-driven platforms
  • Separating AI inference into dedicated services improves maintainability and experimentation speed
  • Geospatial visualization significantly enhances interpretability of public health data

Credits

  • Company: Indihealth
  • Project type: Internship / Internal Research Platform
  • Role: Full-Stack Web Developer