Geocoded conflict-system records condensed into map-ready research layers.
Evidence systems room
AI is useful when it serves sources, maps, and review.
This page frames the technical side of the site: acquisition scripts, public datasets, normalization, geocoding, anomaly review, layer design, and interactive interfaces. The emphasis is research infrastructure that lets complex evidence become legible.
UNHCR public statistics rendered as origin, asylum, and IDP layers.
USGS, NASA EONET, and GDACS records normalized into a risk atlas.
A reproducible local pipeline refreshes conflict, displacement, and disaster datasets.
Research stack
From public sources to visible command rooms.
The working stack now includes genocide and heritage review consoles, global conflict ingestion, displacement and disaster-risk pipelines, COVID regional pages, route optimization, complexity visualizations, and controlled interpretive tools for Linear A. AI belongs in this stack as a disciplined assistant for extraction, classification, and review queues.
Download, parse, and normalize public datasets into stable local JSON and CSV assets.
Separate confirmed records, contextual material, coordinate review, and source-only references.
Expose the work through maps, timelines, filters, record panels, and mobile-friendly atlas pages.
Enter the systems