Business Problem
- Business users did not trust the data from the legacy data warehouse due to lack of transparency into lineage, transformation, and quality
- Data stewardship and rules had recently been implemented, but the data management systems where not integrated with governance platform
- Technical assets lacked logical models for easy business consumption
- Tool were not available to allow stewards to maintain reference data
- Business logic embedded in reports where not understood, cataloged, or managed
- Business users started creating shadow data marts and reports
Lucid Approach
- Developed a 3-phase roadmap to upgrade and expand Collibra DGC with custom workflows, dashboards, asset models, On-the-Go, and integrate it with IBM InfoSphere, IBM Cognos, ServiceNow, MuleSoft, and Tidal.
- Upgraded Collibra DGC and deployed Collibra On-the-Go to the enterprise
- Personalized the DGC user interface with Flagstar’s logo and color themes
- Visibility to data lineage and transformations via integration with IBM DataStage
- Visibility to data quality score via integration with IBM Information Analyzer and custom dashboards
- Automation of data quality issue management through custom workflow trigger via integration with IBM Information Analyzer
- Provided the technical team with Collibra catalog for modeling and change management via integration with IBM Information Governance Catalog
- Enhanced the asset model to organize new integrated data
- Developed the design for using Collibra as the source for reference data management via integration with IBM InfoSphere and custom workflow
- Designed data sharing agreement and concept for managing in Collibra
- Designed solution for advanced lineage using a custom SQL parser
Results
- Improved business trust in data warehouse information via visibility to data quality, transformation, and lineage
- Enabled data steward ownership
- Provided self-service catalog access to enterprise
- Improved time-to-resolution by automating issue resolution workflow
- Made business users less reliant on technical team through understanding of data
- Reduced use of shadow marts