The Client
One of Australia’s four major banking groups, this institution manages a sprawling data landscape across consumer, institutional, and wealth management divisions. As regulatory scrutiny on data provenance, lineage, and AI readiness intensified, the bank required a governance model that could enforce enterprise-wide standards consistently across all divisions, without removing the operational autonomy that divisional teams needed to manage their own data assets effectively.
The Challenge
Fragmented governance frameworks and manual stewardship create a “speed-to-trust” deficit that hampers the organization’s ability to innovate with AI. Without a standardized, interoperable data estate, the business faces significant regulatory risk and misses critical opportunities for scalable automation.
- Fragmented and Non-Standardized Governance: Data governance practices varied widely across divisions, producing inconsistent data flows, duplicated data pipelines, and an uneven compliance posture that increased regulatory risk.
- Manual Governance Processes: Time-consuming manual data governance activities created operational delays and left lineage records outdated, reducing trust in data quality across the enterprise.
- No Interoperability Across Systems: Data assets lacked common context and metadata linkage, making it impossible to leverage data across divisional boundaries or prepare the estate for AI and automation initiatives.
- Regulatory and Audit Exposure: Without end-to-end lineage and provenance visibility, the bank faced material risk in demonstrating data governance compliance to regulators.
The Impact
- 10M+ Data elements harvested
- 3 Divisions onboarded
- 1,000+ Metadata events published monthly
The Solution
Hoonartek implemented the OneGov Divisional Data Governance Platform, a centralized PaaS deployed on GCP that balances enterprise governance standards with divisional self-management capability. Each division operates within a role-based access framework, enabling teams to define and manage their own Business Glossaries, Reference Data, and Data Quality Rules while remaining aligned to enterprise metadata standards. Read-only access to enterprise reference data ensures consistent terminology and definitions are available across all divisions without duplication.
Automated Data Quality pipelines were generated and executed against divisional datasets, with DQ results consumed from Google Dataplex and profiled within the platform. Metadata extraction was built to support diverse source systems including Teradata, Google BigQuery, Datastage, DBT scripts, and geospatial files. Metadata change events are published to Google Pub/Sub topics, enabling 1,000+ events per month to be consumed by downstream systems in near real time. Semantic Discovery logic links technical physical names to enterprise business terms, and CI/CD implementation using Terraform, Codefresh, and GitHub Actions ensures the platform evolves in line with enterprise engineering standards.
Key Benefits
- Enterprise-Scale Metadata Harvesting: 10M+ data elements harvested across Teradata, BigQuery, Datastage, DBT, and custom file sources, creating the most comprehensive metadata inventory the bank had achieved.
- Divisional Empowerment: Three divisions now self-manage Business Glossaries, Reference Data, and DQ Rules within a governed framework, without requiring central IT intervention for routine governance tasks.
- Audit Readiness: End-to-end data lineage and provenance visibility supports regulatory audit requirements and reduces compliance risk across the enterprise.
- Real-Time Metadata Events: 1,000+ metadata change events published monthly via Pub/Sub enable downstream systems to stay synchronized with governance updates in near real time.
- AI Readiness: A governed, contextualized, and interoperable data estate provides the foundation the bank needs to advance AI and automation adoption across divisional boundaries.