Minutes to seconds: getting alert time back at L1 and L2

Where the time goes

Published estimates put a Level 1 review of a customer screening alert at 15 to 45 minutes, and more than 99% of screening hits close as false positives. Across a year of sanctions, PEP/RCA, and adverse media alerts, that is one of the largest lines in a financial crime operations budget: skilled people spending their days confirming that nothing is there.

The reason is not the decision. The alert arrives with a name, a score, and little else. The analyst opens the screening system, the customer record, the case history, and two or three external sources; compares names, dates of birth, and addresses; looks for prior dispositions; and writes a narrative a QC reviewer can follow. Once that is in front of them, the judgment takes moments. The minutes are spent getting there.

Level 1: seconds to assemble, under a minute to close

AML Resolve does the assembly. When an alert lands, it pulls screening data, customer and case-management records, and approved external sources; matches identities statistically rather than by eye; applies your policy; and drafts a rationale with every finding linked to its source. That takes seconds. The analyst opens a finished case instead of an empty one, confirms the disposition, and moves on. Closing an alert with AML Resolve takes less than a minute, and the disposition and evidence are written back to your case-management system.

For alert classes your institution chooses, closing can be automatic. Your policy defines which classes qualify and how conservative the thresholds are; every automatically closed alert carries the same evidence and rationale as one an analyst closed, and anything outside the approved classes goes to a person. The alerts that never needed a human stop consuming one.

Level 2: faster answers on the hard cases

Escalated alerts are where investigative time goes: complex-entity hits involving corporate structures, beneficial owners, vessels, or aircraft; ambiguous matches; adverse media that needs context. A Level 2 analyst spends much of that time asking questions of the data, such as who controls this entity, what is its link to the customer, and what did we decide last time, and then hunting through systems for the answers.

AML Resolve gives L2 analysts an information-rich, function-rich chat over the case: ask questions of the assembled evidence in plain language, pull in additional records and enrichment, run further checks, and build the resolution narrative from the same place. Every answer is source-linked and every step is logged, so L2 resolutions arrive faster and better documented.

The arithmetic

For illustration only: a team working 2,000 alerts a month at 30 minutes each spends 1,000 analyst hours. At under a minute per prepared alert, the same volume takes about 33 hours, and every alert your policy closes automatically drops out entirely. Cycle time moves the same way: alerts are prepared the moment they arrive, approved classes never wait in a queue, and escalations reach L2 with the groundwork done.

Your alert mix, your systems, and your policy will move those numbers. That is why we don't ask you to take ours.

Proving it: Lab, then pilot

The AML Performance Lab takes a sample of screening alerts your team has already resolved and processes them in batch through AML Resolve. From that run we extract the L1 statistics an operations leader needs, analyst time saved and cycle-time improvement by alert type, without touching your live queues.

The L2 gains show up in a pilot, when your analysts work live escalations with the case chat. A bounded pilot on one alert type and one queue can be set up in about a week, because AML Resolve works with the systems you already run. Success criteria come from the Lab's numbers, human approval gates stay in place, and validation takes four weeks, ending with measured results at both levels and an audit-ready record of every case the pilot touched.

Why start here

AI adoption in compliance is rising faster than measurable returns. Alert resolution is one of the few places where that gap closes quickly: the work is high-volume, the cost per resolved alert is already tracked, and the difference between minutes and seconds shows up in the first month. Silk River offers the AML Performance Lab to qualified institutions so you can see those numbers on your own alerts first.

Ready to see the numbers on your own alerts? Request an AML Performance Lab consultation.