Anteact

01/Method

How an engagement runs.

We build every engagement by hand, for one client, one plan, one question. There is no platform and no self-serve path. The work is per request and human-led from first contact to the final re-run.

The premise

Confidentiality makes the work possible. Hold it any looser and the work cannot exist.

You hand over operational data under NDA. We hold it to a minimum-necessary standard, model against it in isolation, and return the findings through a channel you control.

We build the model for you alone. We do not reuse it, repurpose it, or show it to anyone else. There is a sequence behind this, run carefully, one client at a time.

/The sequence. Seven movements.

01/Establish trust

Contact and mutual vetting.

We establish who you are. You establish who we are. Access is selective, and we confirm the engagement is defensive and one we can stand behind before any operational detail moves.

Fig. — mutual vettingabstract
clientAnteact

02/Set the terms

Non-disclosure.

Nothing meaningful moves before a mutual NDA is signed. It states what we receive, how we hold it, for how long, and how we destroy it. We settle the terms while everything is still hypothetical.

Fig. — retention lifecycleabstract
intakeholddestroydefined in agreement

03/Hand over

Secure data intake.

You send only what the model needs, through an encrypted channel. We reduce or abstract identifying detail wherever the rehearsal allows. We hold less, because less held is less exposed.

Fig. — minimum-necessary intakeabstract
retained

04/Rehearse

Modelling and rehearsal.

We rebuild the operational area as a simulation and populate it with thousands to tens of thousands of independent agents: crowds, staff and, where engaged, adversary agents, each deciding for itself across the physical geometry. We run your plan inside it against dozens of perturbations: incidents, surges, comms loss, diversions.

Fig. — model layersabstract
behaviouralphysical

05/Report

The confidential report.

You receive a vulnerability-and-timing report through a secure channel. It states where the principal is exposed and for how long, where the single points of failure sit, and where the route, the message or the comms break under load.

Fig. — dependency graphabstract
single point

06/Review

Review session.

We walk the findings through with you and your security lead in private. The report answers to their judgement. It supports their decisions and stays out of their way.

Fig. — secure deliveryabstract
Anteactclientscoped · encrypted

07/Iterate

Re-run.

You edit the plan. We run it again. Each pass closes the gap between the plan as written and the plan as it behaves. The last run is the one you want to be quiet.

Fig. — perturbationsabstract
planoutcomes

Make contact

The first run is a conversation.

If the work fits, we start with a short private exchange and an NDA. From there the sequence is yours, run as many times as the plan needs.