Continuous monitoring that escalates only on hits

You choose how often each monitored entity is re-screened — daily, weekly or monthly — and you set it per entity rather than for the whole portfolio. A high-risk counterparty or a name on the edge of a designation can run daily, while a low-risk supplier runs monthly, so screening frequency tracks the risk of each case instead of a single blanket cadence. The interval is what turns a one-time screening into continuous coverage: the entity is re-checked automatically on its schedule against the current state of every source, with no manual re-entry, so a designation added after onboarding is caught on the next cycle rather than the next time someone happens to look.

The reason monitoring does not bury analysts in noise is change detection. Each time an entity is re-screened, the platform computes a fingerprint of the result and compares it to the previous one. If nothing has changed, it stays silent — no alert, no ticket, no cleared-again busywork. When the fingerprint differs, it raises an alert, because something that matters actually moved: a new match appeared, a designation was added or removed, an identifier changed, or an adverse finding surfaced. This is the mechanism that makes an alert meaningful: it means the entity's outcome is different from the last time you looked, which is precisely when a compliance team needs to review, rather than a periodic reminder that the same known result is still the same.

Monitoring is designed to run deterministically at scale and involve AI only where it earns its cost. You can monitor thousands of entities with the deterministic Quick Check, which sends no data to AI models at all, and reserve the AI-Enhanced and AI Review levels for the handful of entities that a change actually flags. When change detection raises an alert, that is the moment to escalate the specific name into an AI Review Check for readable match reasoning and false-positive indicators. Mass monitoring therefore sends zero PII to AI models — the base cycle is pure deterministic screening — while the intelligence is spent only on the cases that changed, keeping both your data exposure and your bill proportional to real risk.

When a monitored entity's result changes, the alert appears in your dashboard, where you can see which entity moved, what changed against the previous cycle, and the current screening result. From there you review the change, decide whether to escalate to a deeper level, and document the outcome as evidence. During the beta, alerts are in-app only: you review changes when you sign in, and email notifications are not yet available. The change record is preserved so the history of a monitored entity — every cycle where its outcome moved — is auditable, which is what a regulator expects from an ongoing-monitoring program rather than a single point-in-time check.