- The bank lacked adequate user behavioral data to predict credit card promotions to users geographically using their service around the world.
- Privacy concerns and sovereignty regulations inhibited the centralization of PII.
- Decreased accuracy in financial projections due to a lack of purchase history and user activity data.
- Devron deployed algorithms at the user level to preserve the privacy of underlying user data.
- Devron trained highly efficient promotional recommendation algorithm to recommend restaurants, events, and credit cards to optimize user purchase behavior.
- Decentralized model training resulted in a 30% increase in customer onboarding and retention from increased service adoption and better product recommendations.
Increased Customer Acquisition
More Efficient Promotions
More Data for Analysis
Reduced Time to Insight