AI & Public Health Intelligence

Subject: AI and Public Health Intelligence

Researcher(s): Khahlil Louisy, Alexander Radunsky, Stephanie Buongiorno, Bayan Konirbayev

Institutions: PII, Data-Smart City Solutions (Harvard University), City of Cleveland, OH

The AI & Public Health Intelligence initiative advances a new paradigm in population health surveillance by integrating non-traditional and non-clinical data streams to detect emerging threats at the most granular scales of urban life. Conventional reporting systems often lag behind the onset of crises; our work interrogates how signals embedded in heterogeneous sources, ranging from IoT sensors, home devices, and mobility traces to consumer purchasing behavior and citizen-contributed data, can be algorithmically merged to produce actionable intelligence for public health practitioners.

At the core of this research is a dual agenda: first, to rigorously evaluate which unconventional datasets yield valid epidemiological markers when analyzed in concert; second, to develop methods for multi-modal data fusion capable of resolving risk patterns not only at the city or neighborhood level, but down to the street-level where interventions must often be targeted. For instance, correlating city-level temporal ecommerce data with zoning regulations and restrictions could unearth pockets of heightened cardiometabolic and respiratory illness. Similarly, spikes in smart-thermometer readings with localized air quality fluctuations, pharmacy transactions, and absenteeism records may allow for the identification of influenza clusters days or weeks before clinical systems register the signal.

This line of inquiry is particularly critical in contexts where official health data are delayed, incomplete, or politically constrained, yet where practitioners and frontline responders require near-real-time situational awareness. By advancing methodological frameworks for extracting public health intelligence from the latent structure of everyday data, this initiative seeks to reconfigure the temporal and spatial horizons of epidemic detection, enabling earlier, more precise, and more equitable interventions in urban health.