AML teams aren’t short on alerts—they’re drowning in them, with rules-based systems generating far more noise than signal. This session shows how SAS is tackling that problem with a machine learning layer that sits on top of existing monitoring, learning from historical investigations to rank alerts by risk. Instead of replacing rules or automating decisions, the model helps investigators focus on what matters most—pushing the highest-risk cases to the top while keeping humans firmly in control. The approach combines data quality checks, feature engineering, and model tournaments into a repeatable pipeline that turns messy alert streams into something far more actionable.
Watch the recording
