PrediAI: AI-Powered Diabetes Early Detection App
An AI-powered mobile application for early diabetes risk detection through nail and tongue image analysis, selected as a finalist at GEMASTIK XVIII.
Understanding the Context
Overview
The Problem
Diabetes cases in Indonesia continue to rise, while access to early screening remains limited, costly, and often invasive, especially in remote areas.
The Purpose
PrediAI aims to provide an accessible, non-invasive, and affordable early screening tool using AI-based image analysis to support diabetes prevention and awareness.
My Contributions
Responsibilities
Led the product development process from ideation to final submission
Authored the full project proposal and scientific paper from scratch
Coordinated cross functional collaboration between AI, mobile, and design teams
Managed project timeline, scope, and deliverables throughout the competition
Handled intellectual property (HKI) documentation and submission process
Produced and directed the product demo and competition presentation video
Technologies Used
Tech Stack
📱 Frontend
⚙️ Backend
🤖 AI/ML
🛠️ Tools
Core Functionality
Features
AI Screening
Detects early diabetes risk through AI analysis of nail and tongue images using deep learning models.
Examination History
Stores and visualizes screening history to help users monitor long-term risk trends.
AI Health Assistant
Provides personalized health insights and guidance using a Gemini-powered AI assistant.
Nearby Health Facilities
Helps users find nearby doctors and healthcare facilities using map-based services.
Lab Result Analysis
Upload lab reports for automatic data extraction and explanation.
Visual Design
UI Showcase
Interface design and user experience highlights
Authentication flow and homepage providing access to core features.
Technical Journey
Challenges
Creating a complete project proposal independently for the first time, from problem definition to technical and business analysis.
Developed strong skills in structured problem framing, proposal writing, and aligning technical solutions with real world needs.
Writing a scientific paper while coordinating ongoing product development.
Learned how to translate technical implementations into formal academic documentation.
Results & Achievements
Impact
Measurable outcomes and project success metrics
Selected as a Finalist at GEMASTIK XVIII
Delivered a complete proposal, academic paper, and working prototype
Achieved high usability with an SUS score of 82.5 (Excellent)
Next Steps
Future Plans
Conduct clinical validation with healthcare institutions to strengthen medical reliability
Expand training datasets to improve AI accuracy across diverse demographics
Add offline preprocessing support to reduce dependency on internet connectivity
Integrate with national healthcare systems or telemedicine platforms
Enhance data privacy and security compliance for real world deployment