Research
for Public Sector.

I build end-to-end spatiotemporal outbreak forecasting systems that support biosurveillance and public health interventions.

DOMAIN: HPAI H5N1 (Avian Influenza)

SCALE: U.S. County × Month

Forecasting Methodology

I design forecasting pipelines that translate disparate data streams into early-warning signals, with an emphasis on rare-event reliability.

FORECAST_ORCHESTRATION

End-to-end systems that move from data ingestion to actionable risk outputs.

FEATURE_ENGINEERING

Temporal gating and lag structure help respect causality and reduce forward leakage.

DECISION_SUPPORT

Risk tiers emphasize modeling constraints because missing weak signals can carry real-world costs.