Philosophy

AOH Sync's behavior follows from a small set of principles about where truth lives, who decides, and what happens on failure.

1. One canonical picture of each person

A person appears in many systems (HR, payroll, Entra, shared mailboxes). AOH Sync collapses those into a single identity and keeps every linked record attached to it. See The identity graph.

2. Backend-authoritative access

Every permission decision is made on the server, on every request. The UI reflects what the backend permits; it never grants access. When the two disagree, the backend wins, so a revoked permission takes effect immediately. See Access control.

3. One source of truth, one writer

Each kind of data has exactly one component that writes it. Identity records, secrets, audit entries, and posture scores each have a single owner. Everything else reads from that owner, so there is always a clear answer to "where did this value come from?"

4. Long work happens in the background

Full syncs and provisioning runs run asynchronously. The UI returns immediately and you poll for status; the work is queued, not skipped.

5. Everything that changes state is recorded

Every mutating action is written to the audit trail with who did it, what changed, and which compliance frameworks care. See Audit and compliance.

6. Risk is measured, not guessed

Anything flagged as risky comes from a defined calculation, not a heuristic. Orphan risk and posture are scored from explicit inputs. See Orphans and risk scoring and Posture and anomalies.