Data-Mesh Value Realisation Director — Commercial Banking

The mandate

After three years of domain workshops, platform licences and catalogue activity, the bank still cannot show whether its data mesh has shortened a credit decision or removed a regulatory reconciliation. Lending teams publish datasets under product labels, yet consumers continue to rebuild borrower, collateral and covenant logic in private pipelines. The executive committee has ring-fenced the next capital release until value and control can be evidenced together.

The named deliverable is a Data-Product Value and Control Dossier covering twelve priority lending products. It will contain consumer contracts, lineage-tested control points, unit costs, reuse evidence, service objectives, owner accountabilities and a defendable recommendation to scale, narrow or retire each product. A reusable measurement model and a sequenced eighteen-month investment case must accompany the dossier.

Milestone one, at the end of week four, is a ledger of programme spend, product inventory and disputed benefits. By week ten, milestone two produces three traced credit journeys and an agreed baseline. Week eighteen closes milestone three with six production-product interventions and observed results. The board dossier, calculation workbook and portfolio decision workshop form milestone four in week twenty-six.

Acceptance requires Finance to reperform the benefit calculations from retained source evidence, Model Risk to trace every critical lending attribute to an accountable control, and two independent consuming teams to use at least four products without reconstructing core logic. The CDO and Business Banking COO will jointly sign the portfolio recommendations; a slide presentation without operating adoption will not qualify as completion.

The client will grant read-only access to programme ledgers, catalogue telemetry, pipeline repositories, issue logs, credit workflow timestamps and regulatory reconciliations. Domain owners will attend fortnightly design sessions, while a finance partner and data-control lead will resolve measurement disputes. Production changes remain with bank engineers, and the executive sponsor will clear access or ownership blocks within three business days.

Why this is external work

Platform, domain and transformation leaders each have reputational capital invested in the current design, making an internal stop-or-scale judgement politically difficult. The bank also lacks a practitioner who has priced reusable data products and tested their controls inside credit operations. An independent evidence trail is required before directors release the next funding tranche.

What you will own

• Reconstruct the full cost of each priority product across engineering labour, platform consumption, control operation, remediation and duplicated consumer work.
• Trace borrower, exposure, collateral, covenant and risk-grade attributes through source, transformation, contract, consumption and regulatory reconciliation points.
• Define product-level adoption, reuse, quality, latency and control measures that cannot be inflated by catalogue registrations or nominal subscriptions.
• Select six production interventions where ownership, contracts or interfaces can be changed safely enough to demonstrate observed operational value during the engagement.
• Reconcile claimed credit-cycle, analyst-effort and control benefits with Finance, separating cashable value, capacity release, risk reduction and unsupported aspiration.
• Recommend which domain products should scale, merge, be rebuilt centrally or cease, with investment, migration and customer consequences made explicit.
• Transfer the evidence model, decision log and quarterly portfolio-review method to named bank owners through a witnessed rehearsal using fresh operating data.

Candidate qualifications

• Directed a federated-data or data-mesh portfolio in a regulated bank and can distinguish genuine product consumption from catalogue or governance theatre.
• Personally linked data-platform expenditure to credit, servicing or regulatory outcomes using calculations that Finance later refreshed without external assistance.
• Designed enforceable data contracts and lineage controls for sensitive lending attributes crossing independently managed domains and technology stacks.
• Closed ownership disputes between business executives, central data functions and engineering teams without dissolving federation into uncontrolled local autonomy.
• Evaluated when a domain product should be retired or recentralised, and can explain the commercial evidence used despite prior programme investment.
• Presented a scale, narrow or stop recommendation to an executive committee where the sponsor’s preferred narrative was not supported by observed adoption.

Non-negotiables

• The proposed engagement director must lead discovery, evidence reviews and the board workshop rather than substituting an offshore assessment team.
• Toronto presence is required three days weekly for the opening month and during production interventions, with monthly travel to Montréal.
• No current implementation resale agreement, platform-vendor incentive or contingent fee may influence the portfolio recommendation.
• All borrower-level evidence must remain inside the bank’s controlled analytics environment and be removed from working views at closure.

1. 49 words maximum. How did you prove that a named data product created reuse rather than shifting integration effort to its consumers?
2. 49 words maximum. Describe one federated product you recommended retiring, including the evidence, stakeholder resistance and eventual disposition.
3. 49 words maximum. Which client inputs would you demand in week one to reconcile a data-mesh benefit claim to Finance?