live · IBM Cloud · ca-tor
v0.2 · open sourcegithub ↗

disease outflow forecaster

Pandexis

Pick a disease and a starting city. Slide R0, masks, travel curbs. Watch a calibrated world map of likely spread, with an AI explainer per region.

Built on IBM Cloud · watsonx.ai · IBM Granite chat + embedding · grounded in 17 peer-reviewed papers.

no sign-in · no tracking · world map loads in <1s

Modelled regions
81
Monte Carlo runs
200+
Calibration metrics
CRPS · log score
Explainer
Granite via watsonx
Press the button. Drag a slider. Watch a continent light up.

Pandexis · Disease Outflow Forecaster

SEIR + gravity mobility + Monte Carlo · IBM Z hackathon prototype
sim: ... ms

Top import hubs (median expected cases)

    Click a country on the map (or in the hub list) to see its forecast curve with 50% / 95% intervals.

    AI Explanation

    Click Explain to generate a plain-English narrative for the current scenario, grounded in the SEIR + gravity outputs.

    Nowcast · particle filter(Funk 2018)

    Paste observed cumulative cases for USA (one day,cases pair per line). Each particle is re-weighted by likelihood of these observations under the SEIR+gravity model.