Automate AML alert triage with Amazon Quick and Snowflake Cortex AI
TL;DR
Amazon demonstrates how to automate AML (anti-money-laundering) alert triage using Amazon Quick Flows, Snowflake Cortex, and the Quick Model Context Protocol (MCP). In Amazon's test environment, automated triage shrank alert investigation time from 30 to 90 minutes down to under 5. Actual results depend on alert complexity and data volume.
Nauti's Take
Nauti sees a real efficiency play here: cutting AML alert triage from an hour to a few minutes would seriously relieve overworked compliance teams. The catch — these models can silently filter out genuine risk cases and depend heavily on clean upstream data, so the "under 5 minutes" only holds under ideal conditions.
High-volume banks are the obvious winners; smaller institutions should fix their data foundation first before plugging this in.
Sources
