With risk data quality as the primary objective, we performed a comprehensive data quality assessment and cleanse of more than 100,000 data points to maximize data quality and minimize negative regulatory impact of bad risk data.

• Designed and implemented hundreds of data quality rules, and then used the rules to identify all data inconsistencies and categorize errors into multiple groups to distinguish different types of errors.
• Performed data profiling to determine the root cause of the deterioration of data quality.
• Remediated more than 100,000 data points by focusing on data standardization, deduplication, matching, and resolution of missing values.
• To prevent new errors from being introduced into the data:
  • Redesigned the data entry form to improve the data collection procedures,
  • Designed additional data rules to ensure quality during data conversion and consolidation, and
  • Designed various controls to validate incoming data transactions.
• To eliminate quality deterioration during data conversion and consolidation projects, implemented procedures to ensure comprehensive data profiling and data quality assessment of the source data before the creation of mapping specifications. Data quality definitions and acceptable quality benchmarks were part of the conversion specifications.
• To ensure sustainability of risk data quality, a data quality assessment program was designed that included creating a blueprint to monitor data quality in “live” databases on an ongoing basis to observe data quality trends, identify new data problems, and check the progress of the data quality improvements.