Engagements

Three ways in, each with an end

We do not sell a services catalogue. There are three engagements, each with a consequential question at its centre, a senior decision-maker who owns it, direct principal involvement, a written boundary and a tangible decision product. Most relationships begin with a diagnosis.

Executive Diagnosis

A focused examination of the problem beneath the visible problem.

For a chief executive, board or executive committee that suspects the presenting problem is a symptom.

Typical questions

Why is the AI programme producing activity without movement?
What is the real constraint on this organisation, and what is merely its shadow?
Should this transformation continue in its current form at all?

What it produces

An executive diagnosis, a constraint map, the framing of the decision that follows, and a recommendation that includes what not to do.

Where it stops

Diagnosis ends when the decision is framed. It does not roll silently into a design or delivery engagement.

Institutional Design

The redesign of the machinery the diagnosis found wanting.

For leadership teams ready to act on a diagnosis: the operating model, decision rights, governance and sequencing that follow.

Typical questions

What would this organisation look like designed around what AI now makes possible?
Where should authority over AI-assisted work actually sit?
In what order can the institution change without stopping?

What it produces

A target operating model, decision and governance architecture, organisational interfaces, and a sequenced transition the institution can execute with its own people.

Where it stops

We design the machinery and sequence the change. Your teams and your delivery partners build it; we do not stand up a parallel delivery organisation.

Strategic Intervention

Direct senior support inside a consequential decision or transition.

For a chief executive, board or transformation leader inside a moment that will not wait: a programme reset, a contested decision, a transition under way.

Typical questions

This programme has consumed two years and significant capital. What now?
The board is being asked to approve something nobody has independently challenged.
Who tells us plainly what we are not seeing?

What it produces

Independent challenge, decision support, programme reset design, leadership alignment, and continuing counsel to the chief executive or the board for as long as the situation warrants.

Where it stops

We advise the people who hold the decision. We do not take the decision, and we do not hold administrative rights in your systems.

Five stages. One thread.

That is what an engagement is. This is what happens inside one.

Whichever of the three it is, an engagement runs through the same five stages. The thread is the point: the decision framework that approved an initiative, the authority under which it acted and the evidence it produced are one record, not three.

Diagnose

Structural and portfolio diagnosis: what the institution can execute, what it should build, and in what order.

Govern

Decision rubrics, enablement councils and routing stood up; authority, audit trail and provenance installed.

Execute

Core modernisation and migration sequencing without stopping the business; AI platform and workflow builds.

Adopt

ADKAR-based change programmes: certified champions, trainers, and AI fluency at every level of the organisation.

Prove

Results measured against your own baseline, in evidence a board or a supervisor can inspect.

How every engagement runs

Those are the stages. These are the terms, and they hold across all three engagements.

The principals do the work. Engagements are led directly by the two of us and remain personally accountable from mobilisation through final handoff. Where focused specialist capacity is required we assemble experienced practitioners around the engagement; accountability is never delegated to a junior consulting team.
A defined outcome, at a fixed fee. We contract to a stated end state with stated completion criteria, priced against a disclosed population and a written set of assumptions. Not open-ended billable hours, and not staff augmentation.
Inside your structure, not beside it. We work within your existing organisation and alongside your captive teams and the partners already delivering for you, rather than standing up a parallel delivery organisation that has to be dismantled afterwards.
Boundaries written down first. What is excluded, what we depend on your teams and your partners for, and where our authority stops are all set out before the work begins, on the first page and not in an appendix.
Why not a larger firm. The honest answer is that a global firm brings bench depth, procurement familiarity and someone to escalate to. What it rarely brings is the two people who sold the work still doing it in month seven. Fifty years of institutional experience are in the room for the whole engagement, which is also why we take few of them at a time.
Evidence, not status. Reconciled numbers, a provenance record of what the AI did and under whose authority, a documented exception register and a runbook. Enough for you to take the decision yourself, and enough to defend it afterwards.

What we will not do

And this is the boundary around all of it.

In full, because it is the part that determines whether we are useful to you.

We do not sell tools, resell licences, or take referral fees from vendors. Our recommendation is not a distribution channel.
We do not hold administrative rights in your live systems, and we do not take the decision to go live. Those sit with you, your captive teams and your delivery partners.
We do not settle a contested reading of your data or your records on our own authority. Where the evidence is ambiguous, it goes back to you with the ambiguity stated rather than resolved quietly.
We do not produce a strategy deck in place of a delivered outcome.
We do not hold scope open. Material change is priced and authorised in writing before it starts.

The engagement is structured so that the difficult questions (the missing documentation, the dependency outside your control, the exclusion nobody wants to name) are governed conditions with owners and dates attached, rather than risks discovered late. What you are left holding is capability your institution owns, not a dependency on us.

If this is the conversation you are having internally, write to either of us directly.

zia@bentinckstreet.com mushter@bentinckstreet.com hello@bentinckstreet.com

We take on a small number of engagements at a time.

If your institution is somewhere between AI ambition and AI evidence, that is the conversation this firm is built for. Write to either of us and we will tell you plainly whether it is work we should be doing.

Begin a conversation