Quentin CasaresData and AI leadership for regulated growth
Advisory

Boards, CROs, CDOs and executives accountable to the FCA, the PRA and BCBS 239

Data and AI governance

Critical data elements, lineage, quality controls and AI governance that hold under audit and supervision.

The numbers in the board pack, the regulatory returns and the AI use cases cannot all be defended, and nobody can say quickly which ones can.

What changes

Three outcomes this practice is built to deliver.

  • 01

    Critical data elements with named owners, lineage and quality standards that hold under external audit

  • 02

    An AI governance framework with intake, risk classification, approval gates and human-in-the-loop controls

  • 03

    Regulatory remediation that closes with the supervisor and the auditor, not only internally

Evidence

Case evidence behind this practice.

Alpha Bank London

Governance across 531 critical data elements and audit evidence

Situation
A bank replacing its core banking platform while under BCBS 239 remediation, with an estate of more than 10,000 data elements and no settled ownership of the ones that mattered.
Intervention
Defined the critical data elements and implemented Microsoft Purview across 531 of them with ownership, lineage and quality standards. Led the Finding 13k remediation with PwC and the Joint Supervisory Team, and founded the Data and AI Governance Working Group.
Measurable result
The Deloitte FY25 audit evidence base closed with no material findings.
Relevance
For any regulated firm that must show a supervisor which numbers can be trusted, and why.

MTC UK

Improving data quality across the probation services estate

Situation
Probation services data feeding operational decisions and reporting to the Director of Probation Services, with quality below what those decisions needed.
Intervention
Led the data, analytics and insight function and its data quality programme.
Measurable result
A 23.7% improvement in data quality across the estate.
Relevance
For public sector and regulated operations where data quality is a service risk, not a reporting nicety.

Capabilities

What this practice covers.

  • Regulatory data confidence programme
  • AI consulting for regulated firms

Next step

Discuss a governance mandate