Design-partner program

Measure it on your own history before it sees a live payment

A small number of institutions will run the first pilots. A pilot is a measurement exercise with a written data boundary, not a deployment, and nothing in it can touch a payment.

SHADOW MODE observing
Payment authorityNONE
Transaction impactNONE
IntelligenceACTIVE
Evidence recordRECORDING
Recommended · appliedBLOCK · ALLOW

Two stages

Replay first. Then observe live.

01

Historical replay

Your historical tokenised events and labels
Cenvoriq replay engine, label-leakage guard
Measured signal quality and detection: precision, recall, alert rate, lead time, candidate clusters
exit replay engine and data deleted; you keep the report
02

Live SHADOW

Mirrored live event stream over a signed API
Cenvoriq intelligence layer in SHADOW
Recommendations, dimensions, graph signals and a verifiable evidence record for your analysts
while payments continue normally: no hold, block, delay, approval or modification
Stage A · what Cenvoriq does
  • Runs a data-quality gate that refuses raw identifiers, naive timestamps and leaking labels
  • Replays historical events on a throw-away engine
  • Scores every event with the real engine
  • Identifies candidate suspicious behaviour
  • Measures against the labels you have, at the time you learned them
  • Produces a partner report and reviews it with your analysts

Candidate clusters are never called confirmed fraud.

Stage B · what Cenvoriq does
  • Observes the mirrored event stream
  • Calculates the risk dimensions
  • Identifies network and integration signals
  • Generates recommendations
  • Records evidence you can verify offline
  • Does not block, hold, delay, approve or modify any payment

That is what makes a first deployment low-risk: the only thing that can go wrong is that the intelligence is not useful, and the scorecard will say so.

Who we look for

Ideal design partners

BanksPSPsPayment aggregatorsFintechsRemittance platformsLendersTelecom and payment networks

Useful characteristics

  • Historical event data that can be tokenised and exported
  • Fraud labels where available, with the time you learned them
  • Existing fraud or risk analysts to review results with
  • Multiple payment rails or integrations
  • Willingness to test in replay and SHADOW, including a result that says the engine did not help

What we will not do

Written into the pilot

  • Ask for names, phone numbers, account numbers, national identifiers, PINs, passwords, one-time codes or your tokenisation key
  • Take any authority over payment execution
  • Claim fraud was prevented; a SHADOW pilot measures recommendations against outcomes
  • Charge a success fee tied to fraud figures
The pilot package
Pilot plan and runbook with approvals and nine stop conditions, onboarding guide, data-processing boundary, security pack and questionnaire starter, scorecard and report template, stream quickstart, hosting options and a penetration-test request for quotation. Shared in full on request before any commitment.

Apply

Tell us who you are and what you would want measured

We reply personally. No customer or transaction data through this form.

Do not submit customer or transaction data through this form. No account numbers, card numbers, phone numbers, names of customers, PINs or one-time codes. Describe the problem; the data boundary for a pilot is agreed separately and in writing.

Your organisation’s domain, so we can verify who we are talking to.
Payment rails
Which fraud costs you most today, in losses, refunds or investigation effort? No case details, no identifiers.
Stage A of a pilot replays your own tokenised history. Tokens are derived with a key you keep.

We use these details only to reply about a design-partner pilot. This website stores nothing itself; a submission is forwarded to our intake inbox, and if that inbox is not connected you are told immediately.