Trust
Risk and safety
AI assists, but policy gates and human approvals decide sensitive outcomes.
How we're different
- Evidence system of record, not a case tracker.
- Immutable timeline and explicit approvals with no silent overrides.
- Examiner-ready binders with linked artifacts and tamper-evident logs.
Trust rule: AI drafts, humans decide for sensitive outcomes.
AI drafting is optional and tenant controlled.
Every narrative must cite evidence IDs.
Approvals required for adverse actions.