Trust Center
Intelligence with an audit trail.
AI is the foundation of Relyon, not a bolt-on. Here's exactly how we keep it trustworthy, and how we handle your data.
How the AI stays auditable
Our methodology.
Dual-model cross-validation
Extractions are run and reconciled across more than one frontier model. Agreement raises confidence; disagreement routes to review, we don't ship a number two models can't agree on.
Confidence scoring & thresholds
Every extracted figure carries a confidence score. Low-confidence values are flagged, not silently passed through.
Provenance you can trace
Each figure is tied back to where it came from on your statement, so any result can be verified line by line, not taken on faith.
Human in the loop
Anything below threshold is reviewed by a person before it reaches you. Automation does the heavy lifting; people own the judgment calls.
Security & data handling
How we handle your statement.
Encryption & PCI-first handling
Sensitive data is encrypted, and we follow PCI-aligned practices for anything we touch. We collect only what's needed to help you.
Least-privilege access
Role-based access controls and an audit log govern who can see what inside the platform.
Built to be reviewed
Dated changelogs, traceable results, and clean records, the platform is designed to stand up to diligence.
Our relationships
Clear about who we are.
Relyon is a retail-sales ISO. We're upfront about how processing works and who backs it.
Relyon is a registered ISO/MSP of Metropolitan Commercial Bank, New York, NY (Member FDIC).