How we work · 04
A governed workforce.
Your environment.
Your data. Your control.
Stand up inside your organisation what we run ourselves — a governed multi-agent workforce for a defined function: code review, incident triage, compliance checking, research synthesis. We map your function to our discipline set and configure the gates for your context.
You hold: the deployed workforce, the review discipline and escalation rules and the human decision boundary as configured.
Why this shape takes longer
You are not buying automation. You are changing who decides what.
A production build changes what your organisation can do. A workforce deployment changes how work gets commissioned, reviewed and stopped — which touches decision rights, and decision rights take longer than code. Eight to sixteen weeks is honest about that; a four-week version of this would be automation with a governance document stapled to it.
The scope discipline is one function. Code review, incident triage, compliance checking, research synthesis — one of those, with a real owner. Organisations frequently want three, and three is how this becomes a programme that never lands.
What we deploy is not a reference architecture from a vendor. It is the operating model this firm runs on, configured for your context: which seats your function needs, what each seat’s written standard says, what the review gate checks, and exactly where the human decision sits.
The boundary, configured explicitly
Agents propose, build and verify. Humans decide anything a reasonable person could disagree about, and anything irreversible, legal, regulatory or safety-critical. Agents do not merge to default branches, deploy to production, delete data, or commit anything that speaks outside the box.
How it runs
Operating model first, then the workforce, then the handbook.
It runs on real work while we are still accountable for it. That phase is where the standards get corrected.
- Duration8–16 weeks, fixed scope
- FeeFixed, quoted in writing before work begins
- Runs onYour environment, your data, your control
- GateEvery agent action logged; every claim carrying an evidence state
- IncludesAI Operating Model Design as part of the scope
Scope one function, with an owner
One function, one accountable owner. The named owner is what makes the deployment land; without one it becomes a pilot nobody closes.
Design the operating model
Decision rights, review gates, the human boundary, escalation routes, funding. This is a service in its own right and it is inside this scope.
Map the function to seats
Which disciplines the function needs and what each written standard must say. Standards short enough to be read and specific enough to be failed against.
Run supervised, on real work
The workforce operates while we remain accountable. Work found wrong is rejected against the written definition of done — and any seat may challenge another’s output, including the senior seat’s.
Hand over with the handbook
How to operate it, change a standard, add a seat, and stop it. Plus the audit trail, so the record answers questions after we have gone.
The two lists that matter
What you hold, and what we will not do.
Both are in the engagement letter before you sign it. The second list is the one worth reading twice — it is where most disappointment in this market actually comes from.
What you hold at the end
- The deployed workforce, in your environment on your data
- The review discipline, and the written standard behind each seat
- Escalation rules and the human decision boundary as configured
- A complete audit trail — every agent action logged and attributable
- An operations handbook, including how to change it and how to stop it
What we won’t do
- Autonomous agents without human decision gates
- Workforces without audit trails
- Deployments without rollback or escalation
- Agents that merge, deploy, or spend without approval
- Opaque systems where you cannot inspect the review
Why we can sell this honestly
We are the reference customer.
Two founders deliver this entire company’s work this way — a senior engineer seat decomposing briefs against a version-controlled standards corpus, nine seats building in parallel, a review gate, an integration gate and a human decision before anything reaches a client. It is published in full, including the parts that are awkward.
Read the delivery model →A smaller first step
Team Enablement installs the discipline in your existing engineers without deploying a workforce. Some organisations should do that first, and we will say so.
Which function would you point it at?
Thirty minutes with a founder. Tell us the function and its owner, and we will tell you how we would decompose it — and whether the timeline is realistic.
Fixed fee · Quoted in writing before we start · NDA available