Maximize Value Across the Entire Customer Lifecycle
Provenir AI for Credit Optimization
Challenge Given a credit portfolio and a set of constraints, what is the lending strategy that yields the best return? For example, which interest rate or credit limit increase should be offered to each customer to maximize expected average portfolio return? Our Solution: Provenir AI Provenir has developed an approach using advanced Machine Learning and Linear Programming to identify which lending strategy maximizes the expected average portfolio return:
Predictive Characteristics: Compound, supervised Machine Learning algorithm that predicts probability of good, probability of customer take-up and expected repayment for each customer in the portfolio.
Segmentation: Segmenting the data based on specific criteria; typically aligned to existing or anticipated client segmentation (i.e. product, credit score, affordability).
Clustering: Unsupervised clustering Machine Learning algorithm to identify patterns of similarity in customer behavior within each segment.
Optimization: An optimization algorithm to maximize return, subjected to constraints, to derive the optimal strategy (i.e. interest rate, optimal credit limit) for each individual cluster of data.
Applications: Our optimization solution empowers organizations to identify which lending strategy (i.e. which interest rate or credit limit increase) should be offered to each customer, maximizing the average portfolio return. Leveraging advanced ML together with Linear Programming, Provenir’s AI solution identifies patterns of similarity in customer behavior to ensure that customers with lower credit risk receive lower interest rates and higher limit increases, and vice versa.
Learn More
Page 1Made with FlippingBook Digital Publishing Software