- The operator wanted a solution to retain the customers and understand the reasons for customer attrition.
- Brought together data for around 40k customers from different sources including their personal info, usage data, billing data, and CRM data
- Built an AI model using 1200 parameters, which could predict the customers who were going to churn
- Identified possible reasons for the churn and methods to prevent them
- Identified potential reasons for customer attrition
- Achieved 60% accuracy in predicting attrition
- Enabled the operator to provide targeted services for the identified customers
- Savings of upto 100+ million USD in revenue