Data Quality Platform Cuts Regulatory Audit Prep from Weeks to Hours for UK Energy Transmission Operator
88% automated match rate between cost and planning data, immediately isolating the records needing human review.
One governed platform gave the regulatory team an auditable, repeatable process ahead of the next price control.
Background
This Tier 1 Energy Operator manages thousands of high voltage assets across England and Wales. Every year, they submit a Regulatory Reporting Pack (RRP) to the Office of Gas and Electricity Markets (Ofgem) covering financial and operational performance across its asset base. The RRP is a mandatory submission and the regulator uses it to assess whether they are delivering and spending in line with their price control allowances.
The process for producing this submission had become unsustainable: it previously relied on manual data extracts, spreadsheet manipulation and subject matter expert knowledge applied in a way that nobody could repeat or audit.
Three problems followed:
Data quality issues showed up too late. Errors were caught deep into the reporting cycle, forcing corrections in both the working spreadsheets and the original source systems. By that point, people had already built on top of the bad data.
The process was slow and expensive. Manual reconciliation between multiple source systems limited how often the cost and volume data could be generated and checked. Each cycle took weeks.
Compliance risk was growing. Reliance on ad hoc manual processes risked breaching commitments to Ofgem and blocked any further automation, including performance reporting and wider data reuse across the organisation.
With tighter regulatory scrutiny expected under the next price control, the team responsible for asset strategy needed to move off of spreadsheets, towards an integrated, auditable solution. A proof of concept had already validated the technical approach and the business case, which needed to move to production.
Solution
WeBuild-AI designed and built a data quality and audit management platform running on Azure. It replaces the previous ad hoc Excel workflows with an auditable process for proposing, comparing and approving changes to RRP datasets.
The architecture has three layers:
Data fabric layer. Instead of manual extracts, data from source systems is made available as governed data products through a data fabric. The application consumes a base data product from this layer and nobody needs to copy and paste CSV files by hand any more. This reduces human error and speeds up the process significantly.
Audit application. A web interface acts as the primary workspace. Users review, edit and validate records directly. Every correction is logged with a reason, a timestamp and the person who made it. Edits write back to a separate audit data product, keeping the base data untouched. Data quality rules flag issues at each stage of the audit so problems are caught before sign-off, not after. This provides an explainable solution for when this customer is interacting with Ofgem.
Workflow and collaboration. The platform supports distinct user roles: strategic leads manage reporting periods and deadlines, team coordinators assign and review work by asset category and engineers review individual records. These groups work through audits together inside the platform, rather than passing spreadsheets around by email. This way of working with data has driven an overarching improvement in data quality, resulting in a higher trust in our customer’s data across the data.
Once cleaned and validated, the audit data product can be exported to existing downstream tools for producing the final cost and volume output, or fed back into the source system of record to correct data at the point of origin.
Impact
| Before | After | |
|---|---|---|
| Data quality checks | Manual spot checks in Excel | Automated, every record, every cycle |
| Audit trail | Over-reliance on email threads and human record-keeping | Every edit logged with reason, person, timestamp |
| Reconciliation | Ad hoc, partial | 87.8% auto-matched, remainder flagged for review |
Why WeBuild-AI
WeBuild-AI worked with the client to define the business case and demonstrate quantifiable value from a bespoke build. Our experience building cloud native platforms for regulated industries, combined with the ability to prototype fast and tighten up for production, is what won the engagement. The client had seen enough generic vendor pitches. They wanted a team that could ship working software in their environment, on their timeline.
We shaped the product iteratively based on real user feedback rather than locking in a fixed spec early. As the regulatory requirements became clearer over time, the platform adapted with them thanks to the WeBuild-AI team’s expertise handling everything from UX design to cloud infrastructure. The initial improvement in data quality has been so vast that we’re now expanding the roll-out of the data quality platform into other functions, to support the wider Net Zero shift in our customer’s business model.
