High-level business designs were fundamentally disconnected from the technical reality of the data estate. This prevented a unified view of how process changes would impact data integrity and vice versa.
.png){kind=small}
Modernising VAT Refund Processing Through Integrated Architecture and Active Governance
High-level business designs were fundamentally disconnected from the technical reality of the data estate. This prevented a unified view of how process changes would impact data integrity and vice versa.
Programme Managers lacked visibility into the "Health" of underlying data. While workflow diagrams existed in static tools, there was no way to know if taxpayer master data or invoice records were accurate enough to support automated approval.
VAT data resided in legacy mainframe systems with no real-time validation. This forced a heavy reliance on manual check-ups and spreadsheet-based reconciliations, leading to significant delays in issuing financial support to taxpayers.
Existing documentation was "passive" capturing a moment in time rather than reflecting the live state of the data. This made impact evaluation and remediation options reactive, often discovering data issues weeks after they had already impacted the transformation timeline.
Without a "Ground Truth" of data quality, moving toward an "Automation-First" model carried high risks of processing incorrect refunds, leading to potential revenue loss or audit findings in a system managing £875.9bn in revenue.
Utilised Data Translators to facilitate workshops, mapping the VAT Refund process in BPMN 2.0. Every process node, such as "Validate Invoice Data," was logically linked in ABACUS to its supporting business capabilities and the "VAT Mainframe" application.
Truth Deployed the Informatica stack to scan the physical data estate, identifying exact database tables for VAT Invoices.
This moved the project from assumptions to "Physical Ground Truth," providing a live view of the data infrastructure.
Rule Deployment Configured Informatica Data Quality (IDQ) to run active checks on critical fields.
These scores were pushed via REST API into a central Business Glossary in AXON.
70-80% Reduction in Manual Check-ups
Automation-first processing enabled by trusted data significantly reduced the need for human intervention in the refund cycle.
50% Faster Identification of Data Blockers
The "Stitch" allowed for proactive remediation, identifying data quality issues at the source rather than during the testing or production phases.
Trusted Automated Ingestion (98% Data Health Score)
Achieved high-reliability scores for taxpayer master data, ensuring automated disbursements were accurate and non-refutable.
Predictive Impact Evaluation and Remediation
The reusable metadata knowledge base allowed the programme to evaluate the impact of changes across the entire estate, reducing audit findings and compliance costs.
Scalable Framework for HMRC Digital Transformation
Established a "Strategic Tooling Operating Model" that is now a reusable blueprint for other tax regimes and financial support streams across the department.
.png)
By integrating business architecture with active data governance, this model transforms reactive tax processing into a predictive, automation-first operation that secures both revenue and taxpayer trust.
.png)
