Inspiration

During disasters, the biggest challenge is not the lack of information — it’s the overload of unstructured information.

Victims send SMS messages, social media posts, and emergency reports describing their situation in plain language. Authorities are left with hundreds or thousands of raw messages that must be manually reviewed and prioritized.

Critical cases — such as medical emergencies or trapped families — can easily be delayed simply because they are buried in text streams.

We were inspired to build a system that transforms chaotic distress signals into structured, geospatial intelligence that helps responders instantly understand:

Where emergencies are concentrated

Which cases are most critical

What requires immediate human intervention

What it does

EDIS converts unstructured distress messages into real-time urgency heatmaps.

Each message is:

Indexed in Elasticsearch

Assigned an urgency score

Tagged with geospatial coordinates

Aggregated into a distress intensity heatmap

Using Elastic Maps, responders can:

Visualize high-urgency clusters

Click individual cases to inspect details

Prioritize life-threatening emergencies

Escalate critical cases to human coordinators

The system turns raw messages into actionable intelligence.

How we built it

We built EDIS entirely using the Elastic Stack:

Created structured indices with geo_point mappings

Stored distress messages with urgency scores and metadata

Used ES|QL to sort and prioritize emergency cases

Built heatmap visualizations in Elastic Maps using sum(urgency)

Integrated dashboard views for situational awareness

Designed a human-in-the-loop escalation model for high-urgency cases

The core idea was to combine semantic prioritization with geospatial aggregation — leveraging Elastic’s native capabilities rather than building an external visualization layer.

Challenges we ran into

Designing meaningful urgency scoring

Converting emotional, unstructured text into a numerical urgency score required careful thought.

We had to define realistic thresholds for escalation.

Index mapping & geo_point configuration

Properly configuring geo fields and ensuring compatibility with Elastic Maps required debugging and iteration.

ES|QL joins and aggregations

Structuring queries correctly and avoiding ambiguous references required multiple refinements.

Heatmap configuration

Ensuring urgency aggregation (sum vs count) accurately reflected severity rather than density alone.

Balancing AI with human control

We intentionally avoided full automation to maintain responsible escalation design.

Working within a limited time frame

As a solo builder, managing architecture, documentation, and visualization within hackathon constraints required disciplined prioritization.

Accomplishments that we're proud of

Successfully built a fully working geospatial distress heatmap using Elastic Maps

Designed an urgency scoring framework aligned with real emergency scenarios

Implemented human-in-the-loop escalation instead of unsafe automation

Built a clean, reproducible architecture using native Elastic tools

Delivered a serious, impact-driven system within hackathon constraints

Most importantly, we created something that addresses a real-world coordination problem.

What we learned

Geospatial aggregation is a powerful tool for crisis prioritization

Proper index design dramatically impacts visualization quality

Clear architecture is more valuable than feature overload

AI systems handling critical domains must retain human oversight

Elastic’s built-in capabilities can replace complex custom frontend builds when used effectively

We also learned that clarity of problem framing is as important as technical implementation.

What’s next for Emergency Distress Intelligence System

EDIS can evolve into a scalable disaster response platform by:

Integrating live SMS gateways and API-based ingestion

Adding NLP-based automatic urgency scoring

Connecting to NGO and hospital capacity databases

Implementing predictive clustering for risk forecasting

Supporting multilingual distress analysis

Deploying city-level or state-level dashboards for emergency authorities

The long-term vision is to transform EDIS into a real-time crisis intelligence platform capable of supporting large-scale disaster coordination.

Built With

• elasticsearch