South AfricaFinance and insurance2024 – ongoingDiscovery Bank
Machine learning next-best-action engine for client personalization
Discovery Bank reported a 40% uplift in customer engagement after building a machine learning next-best-action framework
Problem
Discovery Bank ran on a legacy on-premises data environment that struggled with data volume and speed. That held back its aim of personalizing offers and service for each client.
Approach
The bank moved to the Databricks platform on Azure and built a next-best-action framework on it. The framework uses machine learning models, actuarial models and generative AI, and covers segmentation, pricing, lifetime-value projection, personalization, and risk and fraud detection. Its output feeds client servicing, marketing, pricing, risk and fraud teams. AI-generated communication templates help client-servicing agents.
Results
Takeaway
The bank used one shared set of models to serve marketing, servicing, pricing and fraud teams instead of building separate tools for each. The page does not say how the engagement uplift was measured.
Categories
- Size
- Large enterprise
- Industry
- Finance and insurance
- Business function
- Marketing and salesCustomer service
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and MLGenerative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks describes how Discovery Bank in South Africa replaced a legacy data environment and built a machine learning next-best-action framework for personalization, servicing, pricing and fraud. It reports a 40% uplift in the impact of customer engagement activities, a return on investment of more than 500% and model-building capacity of more than 300 models per day.
Full details are in the original sources.
