
Disaster Orchestrator is an AI-driven crisis management platform designed to improve disaster response speed and efficiency. It combines computer vision, natural language processing, and logistics optimization to support decision-makers during natural disasters such as floods. Our system processes imagery data (uploaded by the user or provided as a URL) to detect the severity of a flood and extract relevant insights. It also aggregates social media reports of people in danger, assigns urgency scores to each report, and prioritizes critical zones. Finally, it uses a logistics optimization agent to generate actionable resource deployment plans, for example, assigning boats or trucks to the most urgent zones with estimated arrival times. The solution is built using FastAPI for the backend, Mistral AI for intelligent image analysis, and a clean React frontend for easy interaction. The platform empowers emergency responders, NGOs, and local authorities to make data-driven, life-saving decisions faster than ever before. With its modular design, Disaster Orchestrator can be extended to include wildfire detection, earthquake damage assessment, and additional real-time data sources, making it a scalable solution for multi-disaster management scenarios.
21 Sep 2025