Alert · reviewed · open
Entity velocity spiked: 1 transfers and $667.41 in 24 hours.
- Detector
- velocity_spike
- Severity
- medium
- Program
- Meridian Remit (simulated)
- Subject
- entity enti_sim_meri_7dj6hf539oc
- Transfer
- acht_sim_meri_3yto6v2s9qr · $336.51 · ach outgoing
- Skoor at alert
- null unscored
- Hard signal
- no
- Policy
- policy-v1
- Opened
- 2026-09-17 19:30Z
- Closed
- —
- Decision clock
- none · due — no clock
- Escalated
- —
Draft narrative
What happened. A velocity_spike alert fired for entity enti_sim_meri_7dj6hf539oc under program Meridian Remit. The alert summary states 1 transfer and $667.41 moved in a 24-hour window, exceeding two times the program's average daily volume.
What the evidence shows. Only one transfer is included in the evidence package: acht_sim_meri_3yto6v2s9qr, an ACH outgoing debit of $336.51 USD, settled, created 2026-07-10. This amount does not match the $667.41 cited in the alert summary, and no second transfer record is provided to account for the difference. The transfer carries no return code and is settled. The single signal driving the alert is velocity.sum_24h_gt_2x_daily_avg with a weight of 10; the alert and transfer are both unscored (band unscored, skoor null, n=1).
What was checked. Entity enti_sim_meri_7dj6hf539oc is VERIFIED, not flagged high risk, not PEP, has no review reasons on file, and was last screened 2026-06-21. Prior dispositions: none on record. Counterparty cpty_sim_meri_7kgtsvji9pv has no country listed. The alert routed to review because the detector is not auto-closable, not because of a hard signal (hard_signal: false).
What is recommended. The entity is verified with a clean screening history and no high-risk or PEP flags, and the single transfer evidence shown is settled with no return code, which supports closing this alert. However, the alert summary cites a total ($667.41) that does not reconcile with the one transfer provided ($336.51), leaving a gap in the evidence a person should confirm before final closure. Given the low severity, low signal weight, and lack of any adverse entity or transfer indicators, this does not rise to a hold or escalation, but the amount discrepancy warrants a brief human check rather than an unqualified close.
- Recommendation
- close
- Confidence
- 0.55
- Model
- claude-sonnet-5
- Drafted
- 2026-09-17 19:30Z
- Rationale
- Entity is VERIFIED, not high risk, not PEP, with no review reasons and a recent screening date.
- Only one low-weight signal (velocity.sum_24h_gt_2x_daily_avg, weight 10) drives the alert; band is unscored due to n=1, indicating thin statistical basis.
- The one transfer provided is settled with no return code, showing no execution or repayment problem.
- The alert summary's $667.41 total does not match the $336.51 transfer amount shown, an unresolved discrepancy that limits full confidence in closing without a person's confirmation.
- No hard signal was triggered; route reason is procedural (detector not auto-closable), not risk-based.
- No prior dispositions exist for context or pattern comparison.
Evidence
{
"n": 1,
"band": "unscored",
"skoor": null,
"signals": [
{
"code": "velocity.sum_24h_gt_2x_daily_avg",
"detail": "24h sum above 2× the program daily average",
"weight": 10
}
],
"routeReason": "detector not auto-closable"
}
Skoor signals
| Signal | Weight | Hard | Detail |
|---|---|---|---|
| velocity.sum_24h_gt_2x_daily_avg | +10 | 24h sum above 2× the program daily average |
Decision
Sign in as an operator to decide. Operator sign-in →
Dispositions
No disposition yet.
Actions
Freeze, request to program, and suspend need a second, distinct approver. Clear hold, cancel, and pause execute on the requester's approval. Executed through the bank's own API with the tenant's sandbox key; a dry run when there is none.
No actions requested.