ArgentinaFinance and insurance2020 – ongoingBanco Hipotecario
Machine learning models for personalized product offers
Banco Hipotecario reported 90% more cross-sell and 35% lower acquisition cost after deploying machine learning models
Problem
Banco Hipotecario ran its analytics on a rigid legacy data warehouse. That made it hard to personalize offers and to use machine learning for upsell and cross-sell.
Approach
The bank moved to a cloud data lake on Azure Databricks and used MLflow to manage the machine learning lifecycle. It deployed its first model within three months and a second about two weeks later. The page does not name the models.
Results
Takeaway
The bank tied its machine learning work to two commercial measures, cross-sell and acquisition cost. The page does not say how either was measured.
Categories
- Industry
- Finance and insurance
- Business function
- Marketing and sales
- Innovation type
- Process innovation
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks describes how Banco Hipotecario in Argentina moved from a legacy data warehouse to a cloud data lake and deployed its first two machine learning models. It reports a 90% increase in cross-sell of new products and a 35% reduction in customer acquisition cost.
Full details are in the original sources.
