Emergent

Inspiration

The idea for Emergent was born after the earthquake in Burma, which personally affected the loved ones of our teammates. In the aftermath, reliable updates were nearly impossible to find - rumours and conflicting headlines spread faster than verified information. That experience showed us how fragile communication becomes during crises. We wanted to create something that could bring clarity when confusion is at its peak - a system that verifies, cross-checks, and organises disaster data so that responders and communities can act faster and safer.

What it does

Emergent is a real-time crisis data aggregator that collects and verifies disaster-related information from global and local news outlets, government agencies, and monitoring networks. It cross-references sources, scores their reliability, extracts locations, and displays verified crisis events on a dashboard or API feed - giving users a trusted picture of what’s happening as it unfolds.

How we built it

We built Emergent with Python 3.12 and FastAPI as the backend framework. feedparser powers our multi-source RSS aggregation. spaCy extracts event types and key entities. geopy converts place names into coordinates. A custom verification engine cross-checks reports across sources and assigns a dynamic confidence score.

Challenges we ran into

Simulating real disaster conditions when no major events were happening live. Designing a cross-verification model that’s both accurate and interpretable. Handling inconsistent or incomplete metadata from RSS feeds. Preventing recursive data updates that caused infinite re-renders in our early tests. Balancing real-time performance with data accuracy and reliability.

Accomplishments that we're proud of

Successfully built a backend that automatically detects and verifies crisis events within minutes. Implemented a confidence scoring model that filters noise and surfaces only trustworthy data. Designed a geocoded dashboard feed that maps events globally. Created a testing mode that replays historical disaster data (e.g., 2023 wildfires and floods) to evaluate performance. Transformed a personal experience into a working prototype with real humanitarian potential.

What we learned

Reliable data in a crisis isn’t just technical - it’s ethical. Verification saves time, and time saves lives.

What's next for Emergent

Integrate real-time alerting and evacuation route systems. Expand data coverage to include social media and satellite feeds. Build a human review dashboard for low-confidence events. Launch WebSocket live updates for emergency operation centres. Partner with NGOs, newsrooms, and local governments to turn Emergent into a deployable public tool for crisis response.