AI enablement & integration
Assessment

AI enablement · Sydney and nationally

AI you can actually run.

We work out where AI genuinely belongs in your business, build it into the systems you already use, and stay until your team owns it, on a platform of AI-native tools we built and run ourselves.

Assess your AI readiness See the platform

Free · about 10 minutes · your report is yours to keep

Product stories · real businesses, real constraints

See the problem before you see the technology.

Case studies are most useful when you can recognise the business problem. Start with the moment the old way of working could no longer carry the business—and follow what changed.

Explore all product stories

Pacific Merchandising · business continuity

What happens when the person who knows every customer is suddenly gone?

A much-loved team member with more than 30 years of experience passed away. The business still had to serve its customers—but their history lived across his emails, paperwork and conversations.

“The first order of business was to know the customer, as he had done. But how?”
Read the Pacific story
Customer memory

Long-term customer

One relationship.
Its history intact.
Orders
What they bought
Context
Why it mattered
Next
Where to begin
01 · Business continuityRecover the customer memory

Turn scattered history into a record the remaining team can inspect, correct and use.

Read this story →
02 · Expert knowledgeMake the exceptions teachable

Capture a reviewer’s corrections as operating rules without pretending the system is the expert.

Read this story →
03 · Digital productBuild what pages cannot carry

Connect identity, payments, evidence and a growing catalogue in a product people can trust.

Read this story →

Enablement is the part everyone skips.

Most AI work fails somewhere between the pilot that impressed everyone and the Tuesday morning nobody opens it. Enablement is the discipline of closing that gap, and it is the whole job, not the last ten percent of it.

So we scope it as three stages, quoted before we start, with a decision point at the end of each. You can stop at any one of them and still be better off than when you began.

STAGE 01

Assess

Where AI would actually pay, where it would not, and what has to be true first. Comes out as a written blueprint you own, usable whether you work with us or with anyone else.

STAGE 02

Build

Integration into the systems you already run, not a new island beside them. Fixed scope, fixed quote, working software rather than a slide about working software.

STAGE 03

Adopt

The stage that decides whether any of it mattered. Training, measurement, and the unglamorous work of changing how a team does a thing it has done the same way for nine years.

We did not start with a deck. We started by building the tools.

Six AI-native systems, built in-house and running our own business before they ever ran anyone else's. That is the difference between a consultancy that recommends AI and one that has had to make it work under its own name.

TOOL 01

Customer & revenue

Agentic CRM Live

Low effort to adopt High impact

Contacts, deals, campaigns and works orders, with AI agents doing the parts people put off: enrichment, follow-up drafting, pipeline hygiene. Multi-tenant, so a consultancy can run its own clients inside it.

  • deals & pipeline
  • campaign automation
  • works orders
  • third-party sync
TOOL 02

Publishing

AI-assisted CMS Live

Low effort to adopt High impact

Pages and articles authored with AI drafting inside the editor, live preview, and an edit mode that lets a non-technical owner change their own site without a developer or a ticket.

  • live preview
  • inline edit mode
  • structured content
  • multi-site
TOOL 03

Search visibility

SEO audit engine Live

Low effort to adopt Moderate impact

Crawls a site, produces findings rather than a score nobody can act on, and turns each finding into a prioritised action with an owner. The point is the queue of work, not the dashboard.

  • automated audits
  • findings to actions
  • per-client workspaces
TOOL 04

Answer engines

AEO readiness pipeline Live

Moderate effort to adopt High impact

Scores how visible a business is to AI answer engines, the surface that increasingly sits between a customer and a search result, and logs every run to an append-only ledger, so improvement is provable rather than asserted.

  • answer-engine scoring
  • append-only ledger
  • baseline & re-test
TOOL 05

Knowledge

Folio document workspace Live

Low effort to adopt Moderate impact

Where an engagement’s thinking lives: briefs, reviews, research, decisions, organised per client and readable by the AI agents that need context to be useful rather than generic.

  • per-client structure
  • agent-readable
  • versioned documents
TOOL 06

Onboarding

Client context ledger In design

Higher effort to adopt High impact

A provenance-tracked record of what a business has told us: who said it, when, and how confident we are, so an AI system is built on verified facts instead of assumptions. Under adversarial review; not yet in front of clients.

  • provenance per fact
  • conflict detection
  • signed blueprint

Status labels are honest, including the one that says “in design”. A tool we are still building is not a tool we will describe as finished. The effort and impact ratings are a first pass and need Michael’s and Nick’s sign-off before this is shown to anyone outside the business. Six tools all rated high impact would be worth less than no ratings at all.

Where we do the work.

Fixed scope, quoted before we start. We do not leave until adoption is real, not just reported.

01

AI readiness & strategy

Explore

Not sure which of the below is right for you? Start here. We assess where you are, what AI would be worth in your specific business, and give you a plan you can act on, with us or without us.

02

Custom AI agents

Explore

An agent built for how your business actually works, not a generic assistant your team has to adapt to. Document triage, client intake, quoting, internal knowledge: the jobs where the answer needs your context to be right.

03

System integration

Explore

Most businesses do not need another system. They need the five they already pay for to talk to each other. Quoting to scheduling to invoicing, inbox to practice system, CRM to everything else.

04

Process automation

Explore

The repetitive work eating your team's week: re-keying, approvals, report generation, handoffs between systems that were never introduced. We map the friction first, then automate what survives the mapping.

05

Search & answer-engine visibility

Explore

Being findable by search engines is now half the job. The other half is being citable by the AI answering the question instead. We run both against your site with the same tools we run against our own.

06

White-label platform

Explore

For consultants and agencies who want to sell AI delivery without building the platform underneath it. Our stack, your brand, your clients, multi-tenant from the first line of code rather than retrofitted.

What the first ninety days actually looks like.

Every one of these six steps has a point where you can stop and still be better off than when you started. That is not a sales position, it is how the work is scoped: three stages, quoted separately, with a decision at the end of each.

Stage: Assess

Discovery

Week 1–2

What happens

  • Two workshops with the people who do the work, not only the people who own it
  • We map how the process actually runs, including the workarounds nobody documents
  • Systems audit: what you pay for, what you use, what talks to what

What you get

  • A written map of the process as it is
  • A shortlist of where AI would pay and where it would not
  • The list of things that have to be true first

What we need from you

  • Roughly six hours from three or four people
  • Read access to the systems in scope
  • Someone who can say yes

Decision point. You can stop here. The map is yours and it is useful to any firm you take it to.

Step 1 of 6

We run the measurement on ourselves first.

Three numbers about us, not about a client we cannot name. Every one is checkable, and one of them is deliberately unflattering.

34.1/100 our own AEO score, 16 Aug 2026

Band: low. That is our answer-engine readiness at baseline, measured with the tool we sell and written to an append-only ledger so we cannot quietly restate it later. A firm that will not publish its own starting number should not be trusted with yours.

6 tools built in-house

CRM, CMS, SEO engine, AEO pipeline, document workspace, and the onboarding ledger still in design. We use all of them daily; that is where the defects surface.

2 founders, both in the room

There is no account manager between you and the people doing the work, and no junior learning on your engagement.

What still requires approval

The product stories above are real working drafts. Where a story names an external client, uses a figure or includes personal context, it stays marked as a preview until the client and owner approve it.

That distinction matters. We would rather show exactly what is still being checked than turn a real engagement into an anonymous success story with the difficult parts removed.

Our people.

Two founders, both in the room, for the whole of your engagement. That caps how many can run at once, and it is why we turn work down rather than queue it.

Michael McKerlie

Co-founder · Tech Apostle

Microsoft and Anthropic ecosystem specialist. Personally reviews every enquiry and stays on every engagement until adoption is real.

Nick Hudson

Co-founder · Design and Deliver

Delivery and client outcomes. The person ensuring the work that ships is the work that works, and that your team can own it afterward.

Questions we actually get asked.

Answered as we would answer them on a call, including the two that do not flatter us.

How do we know AI is worth doing in our business at all?

You do not, yet, and neither do we. That is what stage one is for. The assessment gives you a written view of where AI would pay and where it would not, and it is genuinely usable whether you work with us or with anyone else.

We have told clients not to build things. It is a shorter conversation than the alternative.

What does it cost, and how is it quoted?

Three stages, each quoted separately and fixed before it starts, with a decision point at the end of each. You are never signing up for an open-ended engagement, and you are never quoted for stage three before stage two has shown you anything.

There is no hourly rate and no variation for changes that come out of feedback we asked for.

How long before we see something actually working?

Working software in your environment by about week ten on a typical first engagement, with something visible every fortnight from week seven. Before that you get the map and the blueprint, which are useful on their own.

Do we have to replace the systems we already use?

Almost never, and we will argue against it. Most businesses do not need another system — they need the five they already pay for to talk to each other.

If we do think something has to be replaced, it will be in the blueprint with a cost against it, before you have spent anything on building.

What happens to our data?

It stays in systems you control. We work inside your environment rather than moving your data into ours, and access decisions are yours and reversible.

This is a prototype page and the full data-handling position belongs in a written policy that does not exist yet. We are not going to summarise a policy we have not published.

Is this going to cost people their jobs?

That is not the work we are hired for. Savings, where they arrive, arrive as a consequence of people being able to do their existing work better, never as a headcount line.

In practice the processes worth automating are the ones nobody wanted to do — re-keying, chasing, reformatting. If a client's actual goal is headcount reduction, we are the wrong firm and we will say so early.

What if we start and it is not working?

Stop. There are three points where stopping is a normal outcome rather than a failed project, and they exist because a fixed-scope stage that has not proved itself should not roll into the next one.

You keep everything produced up to that point.

Why are some case studies marked as previews?

Some stories contain a client’s name, figures or personal context that still require the client’s approval before publication.

We show that status plainly rather than disguise an unapproved story as a finished case study.

You have questions. We have honest answers.

No pitch, no pressure. Thirty minutes with a founder. We will tell you plainly whether AI is the right move for your business right now, and what it would actually look like.

Start the assessment michael@samurai.rocks