Sentinel
// Methodology

A methodology you can hand to your examiner

Sentinel scores risk the way examiners expect it to be scored, and documents every step so you can defend the result. This page summarizes the model. The full specification ships inside every exam binder.

FFIEC risk categories

The assessment is organized around the risk categories your examiners already use: products and services, customers and entities, geographies, and delivery channels. Nothing exotic, nothing renamed for marketing reasons.

Inherent risk scoring

Each category is built from individual factors drawn from your own data. Every factor has a defined source field, a defined threshold table, and a defined weight. Change the input and the score changes the same way, every time.

Controls and residual risk

Mitigating controls are assessed against each category using a Wolfsberg-aligned model. Residual risk is then derived from inherent risk and control effectiveness using EU AMLA-style residual scoring, so the path from raw data to final rating is fully traceable.

Documented and reproducible

The methodology chapter of your exam binder lists every factor, threshold and formula in the model. Given the same extract, the model produces the same assessment. That is what we mean by glass-box.

// Scoring_pipeline

From data to rating

1

Factors

Raw data points from your extract, each mapped to a source field.

2

Category scores

Factors are weighted and combined into an inherent score per FFIEC category.

3

Controls overlay

Control effectiveness ratings are applied against each category.

4

Residual rating

A residual score per category and for the institution as a whole.

Want to see the model on your own data?

We will walk you through a sample exam binder and the full factor catalog.

Contact