A Social Housing provider urgently needed to predict whether a resident was going to leave their home through distress.
Once a resident had taken this step, they were at great risk of becoming homeless.
With our support, the Housing provider used an AI algorithm to read all the records for every resident, and identify those at greatest risk of abandoning their home – in time for help to be given.
It could detect faint signals that indicated similarities with others who had already abandoned their residence.
This required the ability to find patterns in scarce data, ways to create additional data, the means to employ the patterns found within a prediction model that people across the business could use and understand.