How Pat is governed.

Pat is Patalign's AI assistant. It helps with questions, reminders, and reports. Here is exactly how we keep it in bounds — these are system controls, not policies.

Last updated: July 28, 2026Trust center

A person approves what matters.

Pat drafts; designated humans review and approve customer-facing messages before they send. Approval requirements loosen only with a track record, never by default.

Everything is logged.

Every Pat action is recorded in an audit log — what ran, what it read, what it produced, and who approved it.

Spending is capped.

Every Pat run has hard cost and step limits. Runaway behavior stops itself.

A named person can stop it.

A designated owner can pause any Pat function immediately, platform-wide. Owner on file: C. Garrett.

Pat says what it is.

Every Pat-drafted message is labeled as AI-drafted and human-reviewed. Pat never poses as a person.

Your data stays small.

Pat only sees the minimum data needed for the task at hand, and client assessment answers are never used to train third-party AI models.

Versions are pinned and tested.

Pat runs on specific, tested AI model versions. Upgrades happen only after our regression checks pass — never silently.

When something goes wrong, there's a procedure.

Detect, pause, notify affected customers, fix, and record the post-mortem in the audit log.

Standards alignment.

Our approach is aligned with the NIST AI Risk Management Framework. The math behind every number Pat cites is public on our Methodology page.