QualittiSector: Energy · Distribution

AI for default prediction and predictive debt collection

Machine learning models that predict default risk and tailor the tone, channel and timing of debt collection.

The problem

Defaults had a significant impact on cash flow and created financial risk.
Debt collection was reactive, standardised and inefficient, with no differentiation by customer profile.

Machine learning models based on customer data, consumption and external factors.

Tone, channel and timing of contact adapted to each customer profile.

A significant fall in the share of customers in arrears.

Greater success in recovering overdue debt.

Collection focused on the highest-risk customers, reducing operating expenses.

No inappropriate or excessive reminders for customers who pay on time.