Live Aid

Our team was deeply moved by the ongoing global displacement crises affecting millions of people in regions like Gaza, Ukraine, Sudan, and beyond. While news coverage often focuses on the immediate humanitarian concerns, we recognized a critical gap in understanding the long-term economic impact of these conflicts. Inspired by the original Live Aid concert that mobilized global support for famine relief, we wanted to create a modern technological solution that brings transparency and data-driven insights to humanitarian aid efforts.

What it does

Live Aid is a comprehensive financial impact dashboard that visualizes the economic consequences of displacement in conflict regions. Key features include:

1) Interactive Map Visualization: A color-coded world map highlighting conflict zones with severity indicators, allowing users to quickly identify the most affected regions. 2) Economic Loss Estimation: Leverages open data APIs to calculate estimated economic loss per household in conflict regions, providing a tangible metric for understanding the scale of devastation. 3) Predictive Modeling: Uses machine learning algorithms to forecast future economic trends in affected areas, helping organizations prepare for evolving needs. 4) Resource Allocation Recommendations: AI-powered suggestions for optimal distribution of aid resources based on the severity of impact and specific regional needs. 5) Donor Management: Allows donors to track their contributions of money, volunteers, food, and supplies, creating accountability and transparency in the aid process. Real-time News Integration: Incorporates news events as pins on the map, connecting economic data with on-the-ground developments.

How we built it

Frontend: Vly.ai [Sponored] Next.js App Router for a responsive, server-side rendered application Tailwind CSS with Shad CN for elegant, consistent styling React Query for efficient data fetching and state management

Backend: Convex for databases ML/AI: 1) Wolfram, Neutral Network, and Regularization for predictive modeling on Urgency_Score

2) API: Open AI [Sponored] /Newsdata.io 3) Workflow automation: Orkes-News Scraping-Open AI Anaylsis-Event_Severity_Score

The application follows a microservices architecture, with clear separation between the frontend and backend. This allows for independent scaling and maintenance of each component.

Challenges we ran into

1) Authentication and using Convex for databases 2) Understanding Orkes and Vly.ai workflow 3) Figuring out Auth0 4) Improvement and refining everything

Accomplishments that we're proud of

1) Data Analytics aspect(Model, Scraping, Orkes) 2) Complete Web with various functionality (Visualization/User-Centric Design)

What we learned

Interdisciplinary Collaboration: We learned how to effectively combine expertise in data science, frontend development, backend engineering, and humanitarian affairs to create a cohesive solution. Flexible Architecture: The migration between backend technologies reinforced the importance of designing with flexibility in mind, using abstraction layers to minimize disruption during major changes. User-Focused Development: The process of designing for diverse users—from data analysts to field workers to donors—taught us about creating inclusive interfaces that serve multiple stakeholders.

What's next for Live Aid

Expanded Data Sources: Integrate additional data streams, including satellite imagery analysis and social media sentiment, to provide even more comprehensive impact assessments. Mobile Application: Develop a companion mobile app for field workers to input real-time data and access recommendations while on the ground. Community Engagement: Create features allowing affected communities to provide direct feedback on needs and aid effectiveness, ensuring local voices are centered. API for Partners: Develop a public API allowing NGOs and research institutions to integrate our data and predictions into their own systems. Offline Capabilities: Add support for offline data collection and synchronization for teams working in areas with limited connectivity.

Built With

• convex
• mongodb
• node.js
• openai
• python
• scikit-learn
• tailwind