We help organisations build four things
A strategy for agents, a way to deliver them, assurance that stands up afterwards, and the capability to keep doing it without us.
We are a small consultancy with software behind it. The advice is not a slide pack that ends at a recommendation: each line of work produces something that runs — a register, a gate, a measure, an evidence pack — and the platform is what keeps producing it after we leave.
Components running today · the integrated platform is in development
Agents Strategy
Which work should be done by an agent at all, in what order, and what the organisation is accepting when it says yes. Ends with a portfolio, a decision record, and limits the board has actually seen.
Agents Delivery
Seven stages with gates, from a recorded mandate to an approval by someone other than the builder. Ends with agents released on evidence rather than on confidence, and a change that reopens the gate it affects.
Agents Assurance
Every run leaves a record; the records are scored on five pillars and turned into a ranked list of what to fix. Ends with an evidence pack somebody outside the organisation can check.
Agents Capability
Seats, standards, review discipline and the habits that make the rest hold. Ends with your own people doing this, and a maturity level computed from evidence rather than claimed.
What runs today, line by line
Agentry is in development and we will not pretend otherwise. Each line of work is available now as advice; what differs is how much of it the software already keeps for you.
| Line | Available now | In development |
|---|---|---|
| Agents Strategy | The assessment, the two lenses and the decision record, run with your people. | The portfolio screen that keeps candidates, decisions and refusals inside Agentry. |
| Agents Delivery | The seven stages, their gates and the separation rules, applied to your agents. | The lifecycle and gate queue in the console. The rules are written and fixed. |
| Agents Assurance | A governed gateway that records every call, and five-pillar scoring over the evidence. | The integrated record, the signed evidence pack and the ranked list of fixes. |
| Agents Capability | Seats, the review discipline, seeded errors and reviewer agreement. | Rubric versions, the standards register and maturity computed month on month. |
Where a part is in development, the work still happens — the record is kept with you and imported when the screen ships. We will show you which is which on screen before you buy anything.
How an engagement usually starts
Count what you already have
Two to three weeks. Every agent, script and assistant already running, with an owner against each one. Most organisations find more than they expected, and some with no owner at all.
Fix the worst gate first
Not a transformation programme. The single control that would clear the most agents, applied, measured the following month to see whether it worked.
Hand it over
Your seats, your standards, your review rota. We say plainly what still depends on us, and what does not.
We will tell you where we think agents are the wrong answer. That advice is part of the work, not a reason to end it.
See it working
Thirty minutes, on screen. We will show you what runs today, say what is still being built, and answer what we cannot do yet.
Book a working demonstration