Using our unsupervised machine learning models we take smart data from oodles of big data. And unlike other predictive technologies that train models such as deep learning algorithms or neural nets, because we are looking at the behaviour in data that is built in the current cohort you are not subject to black swan events. Additionally since our algorithms are all open box there is no discrimination or bias. Our Pre-Crime models have been built on pure time series information and give highly accurate results on unseen data.
Not only can our crime modelling software be used by the police it can also be used by the ambulance service to ensure resources are deployed to areas where crime that will result in medical attention being needed.
Using arson data we are able to build Fire detection models that can give time series predictions on when and where fires will start.
With a whole new wave of alternative lenders and challenger banks disrupting the banking and lending industries, a new and fairer way to identify if a consumer is a credit risk is required. Basing future outcomes on past historic data is very difficult in non-stationary datasets. By using our unsupervised machine learning models we can increase your pass rates and reduce your defaults. In the Insurance market our models can give you the propensity for claims (house, motor, life etc)
For both bricks and mortar and online retail stores we have lots of experience building predictive models for many retail players. Our CTO was one of the winners of JLAB (John Lewis & Waitrose Lab) in 2017 with his predictive footfall modelling software which could accurately predict the number of people visiting a store without the need for expensive people counting hardware. Our computer vision algorithms have also been used in a number of retail environments, including self-service scales and customising mobile app shopping preferences using prediction algorithms.