Security
Secure by architecture
Payinference is a decision layer, not a money-movement system. Its strongest security guarantees are structural. What the system never touches, it can never lose, leak or misuse.
Hard boundaries
What the system never touches
These are architectural properties of the platform, enforced in code at the API boundary, not policies layered on top.
Never moves or holds funds
Payinference returns instructions. Your PSPs, gateway and rails execute payments. No funds ever pass through or rest in Payinference, so there is nothing to custody and nothing to settle.
No card data, by schema
Request schemas are strict and reject unknown fields at the boundary. PANs, CVVs and customer identity are not accepted, stored or logged, so card data cannot enter the system even by accident.
A pure decision path
Decisions compute from cached policy, cached provider health and cached pricing. No PSP calls, no LLM calls and no blocking writes on the decision path. Nothing external is trusted at decision time.
Minimal telemetry
Operational telemetry works on bucketed amounts and hashed merchant identifiers rather than raw values.
Accountability
Every decision is accountable
A decision layer has to be auditable to be trustworthy. Every instruction Payinference returns is explainable and traceable.
Reason codes on every decision
Each instruction carries structured reason codes explaining why it was chosen. Decisions are explainable by construction, not reconstructed after the fact.
Decision logs and replay
Every decision is recorded with its inputs, matched policy rules and instruction, so any decision can be replayed and reviewed.
Version-controlled policy
Policy changes are versioned; the rules in force for any given decision are a matter of record.
Outcomes joined to decisions
Reported outcomes are joined to the decisions that produced them, so behavior can be audited against real results.
This page describes architectural guarantees of the platform; it does not claim third-party compliance certifications.