Client:
The client is a renowned UK-based Insurance broker & aggregator with access to risk management and employee benefit services. They aim to fill a market gap for a specialist, independent, client-centric, and service-focused insurance adviser.
Business challenges:
- Largely inefficient monitoring of high-volume client base and not-so-up-to-date products churned £10M since 2018
- Needed an early warning indicator of the likelihood of current clients renewing their existing products
- Efficient monitoring of high volume (small to mid) client base lacked
Benefits achieved:
ITC Infotech delivered a 1% improvement in churn, representing an increase in revenue of ~ £720k. We provided the predictions (churn likelihood) using XGB, Random Forest, Logistic Regression, and Decision tree algorithms that helped achieve 82% accuracy with 0.62 ROC-AUC. Moreover, end-to-end MLOPs integration made tracking, training, deploying, and monitoring the models easy.
