AI Office
An AI team for less than one AI hire.
Three plans, published prices, month to month. Or start with a single workflow for a fixed fee and decide afterwards.
Grow without growing headcount.
Most companies add people roughly in line with revenue. For the administrative half of the business, that stopped being necessary. The question is no longer whether the work can be automated — it's who is going to do it, and what that costs compared to hiring.
Hiring is the obvious answer and the expensive one. A capable AI lead is $150,000 or more fully loaded, takes months to find, and is a single person who might leave in a year — taking everything they learned about your business with them.
Plans
Three plans. Published prices.
You should be able to decide whether this is worth a conversation without having one.
Map
$2,500/month
You know AI matters. You do not yet know where to start.
- A named senior partner as your point of contact
- A living AI roadmap for your operation, reviewed quarterly
- Quarterly working sessions with your leadership team
- Vendor advice — what to buy, what to skip, what to build
- A written assessment of where your data is and what it can support
- Email and shared-channel access between sessions
No build work. This tier buys judgement, not engineering.
Well under half a junior analyst, for senior judgement.
Most common
Build
$5,000/month
You know what you want built and have nobody to build it.
- Everything in Map
- Monthly working sessions rather than quarterly
- Production builds — automations, internal tools, reporting, document pipelines
- Ongoing maintenance of everything we ship
- Direct access to the engineers doing the work
Less than one mid-level hire — who would still need managing, tooling and a laptop.
Run
$10,000/month
You want us operating it, not just building it.
- Everything in Build
- Weekly cadence
- We run what we build — monitoring, alerting, operational response
- A prioritised queue across multiple functions
- Governance: access control, audit trails, data and model policy
- An annual operating roadmap reviewed with your board or sponsor
Roughly one senior AI hire, fully loaded — except you get a team, and continuity when someone leaves.
- Commitment
- Month to month
- Notice
- 30 days
- Annual prepay
- 10% discount
- Price
- Fixed for the first 12 months
- Ownership
- You own everything we build — code, prompts, documentation
- Where it runs
- Inside your own Microsoft tenant
Not ready for a retainer
One workflow. Four to six weeks. Fixed fee.
$6,000
credited in full against month one of any retainer started within 30 days
A monthly retainer is an open-ended commitment to a firm you met three weeks ago. This isn't. It's one bounded outcome with a number attached, and if we miss it we keep working.
The guarantee
Before we start, we agree in writing the number the build has to hit. If it does not hit it, we keep working at no additional cost until it does.
- One process, chosen together, that pays for itself on its own
- Working software in production — not a report or a roadmap
- The measured before and after, in writing
- Built in your Microsoft tenant. You own the code.
How we sequence the work
Start where being wrong is cheap
We never make finance the first build. It is the most common way these projects fail, and the sequencing is deliberate — buyers themselves report trusting AI far more for reading documents than for calculating numbers. We build in that order.
01
Start where being wrong is cheap
Document reading, internal reporting, draft generation. High volume, forgiving of error, immediate to verify.
02
Then the operational workflows
Reconciliation, alerting, approvals — the processes with a human still in the loop.
03
Finance-grade last, and only once trust exists
Payment, pricing, coverage, anything a regulator or a client sees. The accuracy bar is far higher and a wrong number early destroys everything after it.
Your data
Your data never leaves your tenant.
We do not hold your data, because we do not build on our infrastructure. We build inside yours.
Every one of those is implemented today, not aspirational. For most buyers it is a better answer than a certificate, because it describes what actually happens to your data.
- Built inside your own Microsoft tenant
- Entra ID role-based access through security groups
- No static or local accounts
- MFA enforced
- SFTP with IP whitelisting for partner feeds
- Sensitive data masked in every non-production environment
What we will not do
Stated plainly, because a month-to-month contract punishes mismatched expectations immediately.
- We will not replace your ERP or CRM. We build the layer on top.
- We will not start with a finance-accuracy process.
- We will not scope healthcare work touching patient data without a BAA in place.
- We cap intake. When we are full we say so and stop selling.
Will you still be here
One client has renewed us fourteen times since 2020.
A benefits administrator has brought us back for fourteen sequential engagements since 2020 — eligibility reconciliation, invoice automation three separate times as the requirement changed, pre-certification workflows, a data-quality programme — and we still operate their platform today, with alerting, weekly. Every one of those fourteen ended with them asking for the next thing. That is the part worth weighing: fourteen engagements is not one relationship, it is thirteen separate decisions to keep us, each made by someone who could have stopped. Six years is the honest answer to the question behind every retainer decision — will you still be here in year three.
Questions
The things people actually ask
What happens if it does not work?
The Proof Build is one scoped outcome with a number agreed in the SOW. If we miss the target we keep working at no additional cost until we hit it. On the retainers, you can leave with 30 days’ notice — there is nothing to unwind.
Is our data safe?
We build inside your own Microsoft tenant, so your data never reaches our infrastructure. Access is through Entra ID security groups, there are no static accounts, MFA is enforced, and sensitive data is masked in every non-production environment.
Are you going to be here in two years?
We have run one client’s platform through fourteen sequential engagements and we still operate it, with alerting, weekly. That is the most honest answer we can give, and it is a fact rather than a promise.
We tried a pilot and nothing shipped. Why is this different?
Because the first engagement is a fixed fee against one measurable outcome in production, not a discovery phase. If it does not ship, we have not delivered.
Do we need to replace our systems?
No, and we would advise against it. We build the layer on top of what you already run. Replacing a system of record is expensive, slow, and rarely the actual problem.
How much of our time does this take?
On Map, a quarterly session and email between. On Build, monthly working sessions. The heaviest cost is access to whoever actually understands the process — that is the constraint, not calendar time.
Why a retainer and not a project?
Because the second build is always better than the first. The knowledge we accumulate about your operation is what makes each subsequent piece faster — and that only compounds if we stay.
What does this cost compared to hiring?
A capable AI lead is $150,000 or more fully loaded, takes months to find, and is one person who might leave. Map is $30,000 a year, Build $60,000, Run $120,000 — for a team, with continuity.
Twenty minutes.
We ask what your team spends its week on, and tell you honestly whether there is anything here worth doing. If there is not, we will say so.