Deployed terabyte-scale algorithm
Optimized location accuracy
Enhanced customer insights
Refined merchant mapping process
The challenge
Identifying location trends from transaction patterns
A financial institution had insights into card transaction patterns but lacked visibility into customers' home locations and merchant locations. To enhance location-based targeted campaigns, the company sought to develop a scalable algorithm to estimate these locations accurately.
Key challenges
Needed to scale across time, volume, and velocity
Missing customer and merchant location data
High cost and dependency of purchasing data
Required deployment on large amount of data
The solution
AI-driven location intelligence
Enhanced data intelligence
Assessed multiple sources
Added third-party merchant data
Enhanced transactions with locations
AI-powered mapping
Mapped locations via ML
Extracted data with text-mining
Refined for accuracy
Implementation approach
1
Data integration
Scaled framework
Standardized transaction
Integrated data sources
2
Algorithm development
Iterated for accuracy
Validated with sample data
Used ML for location mapping
3
Scalability and deployment
Optimized large amount of data
Supported streamlined processing
Integrated into workflows
The impact
Precision in location mapping
Scalable deployment
Deployed terabyte-scale algorithm
Estimated near to complete customer homes
Mapped almost all merchants
Precision mapping
More than half merchants within 0.3 miles
Almost half customers within a mile
Greater data reliability
Business benefits
Greater data reliability
Improved decision-making support
Enabled targeted marketing
Looking ahead
Future enhancements
Expand model to new markets
Continuous optimization
Refine algorithms for higher accuracy
Scalability and integration
Enhance real-time processing capabilities
