Predicting Tenancy Failure: How AI Can Help Improve Lives- Together Housing.*

Introduction

Together Housing is one of the largest housing associations in the north, managing nearly 39,000 homes and supporting around 80,000 people.

As a housing provider, it is fundamentally a people-focused organisation, but with its ‘Love Our Data’ slogan, Together also appreciates the vital role information analysis plays in delivering excellent services.

One exciting project that combines its people-focused approach with the way it uses data is Together’s tenancy prediction model. This model allows the association to target resources, support residents and help prevent tenancy failures.

How AI is Shaping Tenancy Management

The team at Together Housing worked hard to figure out how to predict tenancy problems before they occur. While it’s impossible to know exactly how long a tenancy will last, their model can estimate the chances of a problem arising, allowing housing management colleagues to intervene and offer help early.

Here’s how it works:

Deep Learning Models: The model uses survival analysis, neural networks and likelihood estimation to understand tenancy patterns.

Accurate Predictions: The model accurately highlights tenancies that might be at risk.

Tenant-Focused: It differentiates between issues that can be solved (like overdue r...