PredPol Predictive Policing Company Founded

PredPol, a predictive policing software company, is founded in Santa Cruz, California by UCLA Professor of Anthropology Jeff Brantingham and mathematician George Mohler. The company emerges from research begun in 2010 when Brantingham recruited UCLA mathematicians to develop algorithms for predicting crime patterns.

The software’s core technology is based on Mohler’s discovery that “self-exciting point processes”—statistical models originally developed to describe earthquake aftershocks—also accurately describe the temporal and geographic distributions of burglaries and other crimes. This research was published in the Journal of the American Statistical Association in 2011 and forms the basis of the ETAS (Epidemic Type Aftershock Sequence) algorithm at the heart of PredPol’s software.

PredPol’s technology ingests historical crime data and generates daily predictions about where and when crimes are most likely to occur, directing police departments to patrol specific geographic areas. The company markets the software as a data-driven approach to reducing crime through optimized patrol deployment.

The founding of PredPol marks the commercialization of predictive policing technology, transforming academic research into a product that will be adopted by dozens of police departments across the United States. However, the technology will later face intense criticism for perpetuating racial bias in policing, as algorithms trained on historically biased arrest data direct disproportionate police presence to predominantly Black and Latino neighborhoods, creating self-fulfilling prophecies of increased arrests in those areas.