How we work

Our engagements are built to leave capability behind, and where you need it, working systems. Here is what to expect.

The four stages

  1. Diagnose (2–4 weeks).

    We interview the people who run the function, review policies, models, systems and reports, and test them against CBU requirements and good practice. You receive a written point of view: what works, what is at risk, and the order to fix it in.

  2. Design.

    We draft policies, methodologies, models, processes and system architecture together with your team, in workshops and working sessions, so the result fits your bank and your people own it.

  3. Implement.

    We stay through go-live: calibrating models, running the first cycles, preparing the first committee packs, and building data pipelines, reports and tools where you ask us to. Where you need a system from a vendor, we write the requirements and help you choose on merit.

  4. Hand over.

    Documentation, training, the code and models we built, and a clear date when your team runs it without us. Optional follow-up reviews at three and six months.

Ways to engage

  • Diagnostic: fixed scope, fixed fee, two to four weeks. The usual starting point.

  • Project: a defined deliverable, such as an IFRS 9 methodology, an anti-money-laundering risk assessment, an ALM framework or an IT strategy.

  • Delivery: a defined build to a fixed scope, such as a data mart, a set of regulatory reports or an AI prototype.

  • Retained advisory: ongoing senior support for a CRO, CFO, CIO, head of compliance or board, a set number of days each month.

Advice and delivery, with conflicts on the table

We advise, and where you want us to, we deliver: data platforms, regulatory reporting, dashboards and AI tools, built by our own team or by implementation partners we name in writing. We keep three rules.

  • We disclose before we propose. If we, a partner, or a company one of our people has an interest in could be paid to build or supply a system, we tell you in writing before any proposal.

  • We do not mark our own homework. If we are a candidate to build a system, we do not also run the selection. Your team, or an independent party, makes the choice.

  • You own what we build. The code, the models and the documentation are yours.