Glublood

Project Overview

Glublood is an advanced mobile health ecosystem designed to prevent, predict, and manage diabetes. Moving beyond simple logging apps like Diafit, Glublood features a highly comprehensive set of features, utilizing robust system architecture, clean design patterns, and hardware integration.

The application connects a cross-platform mobile frontend with a multi-service backend, scraping external listings and processing machine learning models locally.

Advanced Features

• BLE Glucose Meter Integration: Syncs blood sugar levels instantly with external hardware meters over Bluetooth Low Energy (BLE) protocols, removing manual entry errors.
• Diabetes Risk Prediction (C4.5 Algorithm): Employs a custom classification model using the C4.5 decision tree algorithm written in Python to evaluate risk indicators.
• 8,000+ Scraped Food Database: Integrates a custom database of over 8,000 local and international food items. The listings were collected using a custom distributed scraping engine built with Golang and Selenium.
• UPC Barcode Scanning: Integrates Open Food Facts API, allowing users to scan product barcodes to check sugar, carb, and nutrition guidelines instantly.
• Robust Clean Architecture: Structured using industry-standard design patterns to ensure maintainability, testability, and enterprise-grade performance.

Architectural Stack

• Mobile Client: Engineered in React Native for high-performance multi-platform UI renders.
• Core Backend Service: Driven by a Laravel backend service storing structural entities.
• High-Performance Services: Uses a Golang service to compute food lists and scrape items.
• Predictive Engine: Uses Python to compile and evaluate decision tree algorithms.
• Relational Storage: Relies on a highly optimized PostgreSQL schema.