Success Stories > Data Platform
Data Platform for Customer Segmentation & Fraud Detection
Project Background
Perennial helped a card issuing and payment platform from South East Asia to build a data platform for customer segmentation and fraud detection. We leveraged Google cloud services to build a scalable platform for real time detection of fraud. With the help of the below architecture the business was able to gain insights by dividing the customer into segmentation and it also helped with fraud detection. Multiple dashboards were also created for deriving insights from the landing cycle.
The architecture includes:
Solution Flow:
- Bigquery Biglake table to sent to BigQuery business layer
- Data in Dataproc is then sent to vertexAI for machine learning purposes.
- Data from BigQuery business layer is sent for visualization purposes
- The entire architecture is governed using data governance
- Proper IAM and monitoring, altering is applied across the difference services
Key Highlights
- The multi-tenant architecture and config services allow the Client to onboard a new tenant within a week as opposed to other single-tenant platforms which require a couple of months to onboard new tenants
- The white-label mobile apps allow the Client’s potential customers with a faster go-to-market to test their concepts and offer value added services to their existing client base