We pride ourselves on transparency, a commitment to reducing harm associated with over-policing, and a focus on helping officers find the best tactical solutions to improve their communities. In our resources archives you can find in-depth information about HunchLab, including how our models work and what others have said about us.
We believe that collaboration and transparency will move policing forward faster than patent-pending algorithms and black-box software products. As such, we are transparent about what data we use, the theories underpinning our model, and how our algorithms are built and operate. We consistently share HunchLab’s inner-workings with police departments, journalists, and activists.
HunchLab includes features designed to reduce the potential harm caused by over-policing. We use a probabilistic model to vary where officers are sent, and we show officers a 15-minute timer when they enter a mission area to encourage them to patrol without over-saturating that area. Further, for crime types with enforcement biases, we utilize public reports of events instead of police crime reports.
While software can help police departments allocate their resources effectively, software alone does nothing to reduce crime. The tactics that officers employ in the field are what make the difference in improving communities. Advisor can help departments figure out which tactics work best so that they can more easily adopt evidence-based approaches to combating crime.
In this video we give a brief 10 minute overview of HunchLab, a proactive patrol management system that aimed at helping Police Departments design more effective patrols.Recording
In this video, we outline the concepts behind HunchLab and give a brief demo of what the product has to offer. We provide an overview of Predictive Missions and Advisor, highlight some changes with Sidekick, and take a look at the Dashboard.Recording
Check out the backstory and origins of our product by reading our whitepaper about HunchLab 2.0. This document goes over the overall goals of the product, discusses security details, provides information about Azavea, and does a deep dive into how our modeling works with an FAQ.PDF
How can police departments learn what policies and tactics are effective? This webinar features a preview of a new module for HunchLab that assists departments in the discovery of new evidence-based policing strategies.Recording
From pins on a map to forecasted squares, crime forecasting has evolved drastically over time. This short video by The Verge explains the evolution of predictive policing and compares and contrasts some of the current systems being used.Recording
Algorithms are powerful mathematical tools which shape much of our modern life. They are not without dangers, and in this interview with the BBC, the HunchLab team talks about how we address these problems within our software system.Recording
Have you wanted to explore a crime forecasting tool but hesitated because you were not sure how to get started trying such a tool? Eleazer (Lee) D. Hunt, PhD, the Manager of Information Systems at Greensboro PD, is here to share his department’s experience with testing HunchLab in this webinar.Recording
With the rising interest in crime forecasting tools, important questions concerning ethics, privacy and fairness have been raised. We know that there are some misconceptions when it comes to the topic, and we aim to discuss some of the charged rhetoric and beliefs that surround the term.Recording Slides
In this article, St. Louis County trials HunchLab a year after the death of Michael Brown. The article highlights how HunchLab was used operationally and how it impacted St. Louis County.Article
Have you wanted to explore a crime forecasting tool but hesitated because you were unsure of how it could work for a European law enforcement agency? In this webinar our team discusses how HunchLab can support the European market by providing a translated user interface and data residency within the EU.Recording
Given that HunchLab can use many different types of data within its models, how do you prioritize the data to use? This webinar outlines the types of data that HunchLab can use and ways to think about prioritizing data sets.Recording
Is the best allocation of resources always going to the highest risk locations? This introduces the concept of metered dosage to improve HunchLab recommendations.Recording
What can we learn from the variables used to predict high risk? This webinar uses HunchLab to build predictive models across four U.S. cities using the same data sets and then compares the resulting models.Recording