VietnamFinance and insurance2024 – ongoingTPBank
Credit card propensity models built on IBM Cloud Pak for Data
TPBank's credit card propensity model lifted customer conversion by 24%
Problem
TPBank, a Vietnamese digital bank with 12 million customers, wanted to target credit card offers better. It needed to build and deploy machine learning models on its transaction data faster.
Approach
The bank used IBM Cloud Pak for Data, with Watson Studio, Watson Machine Learning and Watson Pipelines. Its data science team followed the CRISP-DM method: data preparation, feature engineering, model development and deployment. Models were trained on up to 25 GB of structured transaction data.
Results
Takeaway
A propensity model gave the bank's card marketing a measurable conversion gain. A shared model platform also let the team deploy more models.
Categories
- Industry
- Finance and insurance
- Business function
- Marketing and sales
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
IBM reports how TPBank built credit card propensity models on Cloud Pak for Data. It gives a 24% conversion result and says the bank deployed 15 models in 2024. It also reports shorter model development and deployment times, which are platform figures.
Full details are in the original sources.
