Intelligent Risk Management: Natural Language Processing Real-Time Triage of Police Calls for Service

The Seattle Police Department (SPD), in collaboration with the National Policing Institute (NPI), proposes the full-scale implementation and process evaluation of an Intelligent Risk Management (IRM) system for 911 call triage. This project aims to enhance the accuracy of risk assessments for incoming emergency calls, facilitating more effective use of alternative responses, such as dispatching co-responders for mental health crises. The IRM system leverages natural language processing and advanced statistical models to classify call severity in real time, reducing the need for an "all-hazards" police response. Through this project, SPD seeks to refine, test, and deploy the IRM system citywide, improving response times, operational efficiency, and community safety. The proposed project will involve rigorous testing to ensure the system's accuracy and integration with existing operations, followed by comprehensive training and full-scale deployment across Seattle's emergency response network. A mixed-methods evaluation conducted by NPI will assess the system's implementation process, operational impact, and potential for replication in other jurisdictions. This initiative addresses the broader trend in US policing toward diversified response models, which aim to reduce the reliance on armed officers for non-violent incidents, particularly those involving behavioral health crises. By optimizing call triage, SPD expects to recover significant police capacity and enhance public trust. The project's outcomes will include a final evaluation report and a guide for national replication, contributing to BJA's objectives of fostering innovative solutions and enhancing community-police relations.