Creation of a robust tool to optimize credit assessment
Objective
A US-based financial services firm wanted to create an end-to-end real-time system for credit portfolio assessment.
Challenges
- 200,000 business rules creating difficulties in monitoring, changing and enriching data pipelines.
- Poor data quality limiting trust in data.
- Challenges handling non-English languages in underlying data.
Approach
- Sourced over 250k reports and processed 2Bn+ fields of data.
- Performed data enrichment, validation and quality assurance.
- Designed, created and maintained databases to store the data.
Impact
- Credit analysts were able to focus on core analytical activities.
- Standardized and optimized processes.
- Achieved a 99.99% data accuracy threshold.
- Reduced time for data sourcing, extraction and integration.
Questions