A practical first step

Choose one vertical. Align the rules. Pilot with a limited cohort.

The objective of a first engagement is not a finished platform. It is written alignment on the control matrix, the pilot architecture, information-security expectations, and commercial terms — inside about ninety days.

What we agree on before anything scales

A sponsoring institution: One bank or direct processor agrees to co-design the program and approve the controlled pilot. We do not build a control framework in isolation and present it as finished.
One vertical: A single initial vertical with real merchant demand and a manageable legal framework. Breadth comes after the controls have been tested against something.
A legal control matrix: Counsel validates the jurisdiction, product, buyer, marketing, documentation, and escalation rules before any of them are written into software.
A technical boundary: The platform enforces eligibility without storing raw cardholder data wherever the integration allows, and without taking custody of settlement funds.
Security readiness: A formal PCI DSS scope assessment and a SOC 2 readiness program, sequenced so the institution's vendor diligence has something to review.
Measurable outcomes: Approvals, interventions, exceptions, losses, audit response time, uptime, and false declines — agreed in advance so the pilot can be judged rather than argued.
Channel separation: Merchant relationships, pricing, volume, and pipeline belonging to a participating ISO are partitioned from any sales organization using the platform, contractually and technically.

What ninety days produces

A pilot that ends with an opinion is a pilot that has to be run again. These are agreed before it starts, so the decision at the end is a reading rather than an argument.

DeliverableOwnerByWhat it settles

Control matrix, signedInstitution counsel with usDay 30What may be sold, to whom, where, on what evidence

Policy version one, approvedInstitutionDay 30The rules the engine will actually run

Integration against existing checkoutUs with the merchant platformDay 45No platform migration, no new acquirer relationship

Cohort boardedUs, reviewed by the institutionDay 60Capped at the agreed number and not exceeded

Evidence package, sampleUsDay 60What an examiner would actually receive

Measured result against baselineJointDay 90Whether the gates pass, on numbers agreed at the start

Measured against a baseline, not against hope

Each metric is captured before the pilot as well as during it. A number with nothing to compare it to is a claim.

Approvals and false declines

Approval rate, and transactions stopped that should not have been.

Interventions and exceptions

How often a human had to act, and under whose authority.

Loss events

Counted, with the rule that did or did not catch each one.

Audit response time

How long a complete evidence package takes to produce.

Exit criteria, both directions

The institution can stop at any point: Boarding halt, program halt and merchant halt are live from day one and belong to the institution. A pilot that can only be stopped by agreement is not a controlled pilot.
Failing a gate ends it: If a required gate does not pass at day ninety, the pilot does not extend by default. It closes or it is re-scoped with the institution, and the cohort does not grow in the meantime.
Passing does not mean portfolio-wide: A passed pilot earns a second program or a larger cohort under the same controls. Nothing switches on across a portfolio as a result of ninety days.

Start with one program

If your institution has a vertical it has considered and set aside for lack of control, that is the conversation. Tell us the category and the reason it was set aside, and we will come back with the control matrix it would take.

Email programs@sentryfox.net