Project Submission – Track 3: The Proactive Public Health Shield

1. Project Overview

MySihat
An AI-powered media-monitoring platform enabling early outbreak and epidemic detection. It shifts officials from crisis reaction to proactive prevention.

---

2. Our Solution

Problem Statement

• Public health teams often work reactively, detecting outbreaks only after they spread.
• Vast unstructured public data (social media, news, community reports) is underutilized.

Proposed Solution

• Use keyword media monitoring to alert officials of potential health outbreaks based on public mentions.
• Combine real-time data ingestion, AI-driven classification, and risk analysis to enable rapid, targeted actions.
• Designed for speed in public health threat monitoring.

Value Proposition (End-User Focus)

• For analysts: Drastically reduce the manual work of sifting through public data, enabling earlier alerts.
• For decision-makers: Better allocate resources to emerging threats.
• For communities: Faster containment, less exposure to misinformation.

Vision

• Become the default national early-warning layer for outbreak detection.
• Expand to global public health monitoring, multi-language, and multi-disease tracking.

---

3. MySihat Dashboard

Design Philosophy

• Visually striking dashboards with clear key metrics always outlined for easy inspection.
• Seamless navigation.
• AI-generated web scraping and risk assessment.

User Journey

• Dashboard: General overview and map information of outbreaks based on keyword media mentions.
• Early Warning: Early warning dashboard for adding detection thresholds. AI-predicted alerts for possible outbreaks.
• Keywords: Adding custom keywords for media monitoring.
• Media Monitor: Media monitoring of social media and news outlets that are used for outbreak detection.

Wow Factor

• Live, AI-powered threat evolution – location/hot spots in disease outbreaks around Malaysia.
• Media mentions – scraping media mentions for real-time data.

---

4. Technology Stack

Core Architecture

• Data Sources: Social media APIs (Twitter/X, Facebook), news RSS feeds, and EXA scraping.
• Processing: LLM pipelines for language detection, topic modeling, sentiment analysis, and media monitoring.
• Visualization: Interactive dashboards using Next.js, React, and OpenStreetMap.
• Cloud Infrastructure: Supabase.

Innovation in Approach

• Combines epidemiological modeling with social media analytics in one platform.
• Proactive misinformation flagging integrated into outbreak tracking.

---

5. Scalability & Next Steps

• Expand data sources to cover regional radio, blogs, and WhatsApp public groups.
• Integrate with national emergency alert systems.
• Deploy as a cloud-based SaaS for multiple countries.