Approach
Start where the work already repeats
Most AI programs stall because they begin with tools instead of tasks. Four phases, in order, each with a defined output you can hold us to.
Each phase is fixed-fee and ends in something you can hold. You can stop after any phase, and what you have built stays yours.
Two weeks
A written readout, a permitted-use position grounded in your real agreements, and a fixed-fee quote for the build if you proceed.
Four to six weeks
Reviewed prompts and tools, versioned and built on your own methods, templates and constraints.
Four to eight weeks
Access, guardrails and training, team by team, in the tools you already use.
Quarterly
A review gate inside your existing QA step, plus versioning, disclosure standards and an evidence trail.
Maybank method · bar length shows typical duration, longest phase eight weeks
Buy after you agree, not before
Nearly every stalled program we see has made the same sequencing mistake: it bought before it agreed. A tool can work in one enthusiastic team and still fail to spread because nobody has agreed what the work is, who owns the standard, or what the client agreements permit. Those are operating decisions, not licensing problems. Doing the work in this order also reduces wasted spend: the assessment eliminates options before you evaluate them, and governance is designed in rather than added as remediation.
{{ p.name }}
{{ p.duration }}
{{ p.lede }}
{{ x.text }}
What happens
- {{ s.text }}
What you get
- {{ o.text }}
Your side
What we need from you
Engagements that go badly usually go badly for one of these reasons, so we are explicit about them before we start.
A named owner
One person with budget authority and a mandate. Not a committee, and not someone doing this alongside a full client load.
Access to your strongest practitioners
Roughly two hours each from the people who deliver the work best. The method comes from them, not from us.
Your executed client agreements
Not the template — the executed agreements with your largest clients. The permitted-use position depends on what was actually signed.
Honesty about shadow usage
People are already using these tools. We need to know where, without anyone being disciplined for saying so.
Which model fits which problem
We are not the right answer to every version of this question.
Five other routes exist and three of them are better than us for particular problems. Knowing which one you are actually facing saves more money than any fee negotiation.
A major strategy or Big Four-style consultancy
When global transformation scale and enterprise program management dominate the problem — many countries, many entities, a change program that needs hundreds of people and a governance structure to match.
Choose them, not us.
A vertical AI software vendor
When the primary need is a mature packaged product for your sector and your delivery model is close enough to the product’s assumptions that configuration is enough.
Choose them, and we can help you decide which.
An AI implementation consultancy
When the strategy and the operating decisions are already settled internally and what remains is build capacity against a defined specification.
Reasonable, if the specification really is settled.
An independent AI advisor
When you need a thinking partner for a bounded question and do not need anyone to implement, govern or reprice anything afterwards.
Often the right first call.
Your own innovation or transformation team
When the firm already has the authority, the capability and the capacity to execute. Authority is the one most often missing.
Choose them, not us.
Maybank
When the problem crosses delivery workflow, governance, implementation and fee economics at once — and when the commercial question of where the efficiency lands has to be answered rather than inherited.
This is the case we are built for.
Maybank view · categories of alternative, not an assessment of any named firm
Common questions
Questions about how this runs
What happens in the first two weeks?
The assessment maps the work, contracts, existing capability and commercial position, then ends in a written scope and fixed-fee quote.
What do you need from us?
Access to how the work is actually produced, one owner with authority, and willingness to decide where efficiency should land.
Do you advise or do the work?
Both, in that order: we rebuild the delivery method with your team, then leave the assets with you.
What happens after deployment?
Quarterly governance keeps the library, review gate and disclosure position aligned with actual usage.
See the evidence behind the sequence
Why the work starts with a service line, not a tool.
The executive briefing shows why the work starts with a service line and its economics rather than with tool selection.
Phase one starts with three minutes
The online assessment is a compressed version of the two-week assessment. It gives you a useful first read without requiring a call.