Solution · Customer Risk Assessment

Customer Risk Assessment That Stands Up to Audit

One Constellation computes Customer Risk Assessment scores using a configurable matrix across customer, product, geography, channel and behavioural risk factors — recalculated dynamically as the relationship evolves, with the methodology and weighting exportable as evidence.

The Challenge

Risk Scores That Are Easy to Compute Are Also Easy to Pick Apart

The risk-based approach is the foundation of every modern AML regime. FATF Recommendation 1 requires it. MAS, FCA, FinCEN, EU 6AMLD and AUSTRAC all build on it. The core obligation is that the depth and frequency of customer due diligence should match the risk the customer relationship presents.

Which means risk has to be calculated — credibly, consistently and defensibly. A score that is just "high / medium / low" with no documented methodology is the first thing a regulator picks apart in an examination. A score that doesn't recalculate when the customer's behaviour changes is a finding waiting to happen. A score derived inconsistently across the customer book is worse than no score at all.

One Constellation handles Customer Risk Assessment as a structured, transparent calculation. Customer, product, geographic, channel and behavioural risk factors are weighted via a configurable matrix. The resulting score is auditable to the factor level. And the assessment recalculates dynamically as the customer's circumstances change — new transactions, jurisdictional updates, sanctions designations, PEP status changes — without manual intervention.

How It Works

Risk Calculation That Compliance Officers and Auditors Both Understand

Multi-factor scoring

Risk is assessed across five standard dimensions: customer profile (PEP, beneficial ownership, occupation), product (which services the customer uses), geography (residence, citizenship, transaction destinations), channel (face-to-face vs remote onboarding) and behaviour (transaction patterns over time).

Dynamic recalculation

Scores aren't snapshot. They recalculate when any underlying factor changes — new sanctions designation, jurisdictional risk elevation, transaction pattern shift, PEP status confirmation. The customer's risk classification stays current automatically.

A defensible risk assessment is one where the auditor can ask 'why is this customer scored medium' and see the exact factor weighting that produced the answer — in seconds, not by reverse-engineering a spreadsheet.
— Design principle, One Constellation risk engine
Operational Impact

Risk Assessment, Measured

Performance figures from One Constellation CRA across production customer books.

5
Standard Risk Dimensions
Auto
Dynamic Recalculation
Per-Factor
Audit Trail Granularity
Configurable
Matrix & Weighting
100%
Methodology Transparency
What's Included

Risk Assessment Across the Customer Lifecycle

CRA on One Constellation covers initial scoring at onboarding, dynamic recalculation through the relationship, periodic review triggering, and audit-ready evidence of every score change.

📊

Onboarding Risk Scoring

Every new customer is risk-scored at onboarding using your configured matrix, with the factor-level breakdown visible to the analyst.

Explore Onboarding →
🔁

Dynamic Recalculation

Scores update automatically as underlying factors change — sanctions, PEP status, geographic risk, transaction patterns — across the entire customer book.

Explore Portal →
📋

Periodic Review Triggering

Risk classification drives review cadence — high-risk customers reviewed annually or more frequently, low-risk customers on a longer cycle, with full audit trail.

Explore AML/CFT →
Capabilities

Risk Assessment Built for the Way Regulators Actually Inspect

The features that turn risk classification from a policy fiction into an operational reality.

🧮

Configurable Risk Matrix

Factor weighting, scoring scales, and risk tiers all configurable per customer book and jurisdiction — the underlying methodology stays consistent, the calibration matches the business.

🌍

Geographic Risk Layer

Country risk scoring based on FATF lists, Basel AML Index, Transparency International CPI and internal jurisdictional risk — updated as classifications change.

👤

Customer Profile Factors

PEP status, beneficial ownership complexity, occupation risk, source of wealth profile, customer history — all integrated as scoring inputs.

📦

Product Risk Weighting

Product or service risk built into the matrix — correspondent banking, trade finance and crypto services weighted higher than basic deposit accounts.

📈

Behavioural Risk Signals

Transaction-monitoring alerts, velocity changes and pattern deviations feed back into the customer's risk score automatically.

📋

Per-Factor Audit Trail

Every score change preserved with the factor that triggered it, the previous and new values, and the timestamp — auditable to a level no spreadsheet matches.

Regulatory Alignment

Risk Assessment Aligned to the Risk-Based Approach Each Regulator Mandates

Every modern AML regime mandates a risk-based approach with a documented, defensible customer risk assessment methodology. One Constellation supports the specific obligations of each.

🌟
FATF Rec 1
Risk-Based Approach Standard
🇸🇬
MAS Notice 626 §6
SG Customer Risk Assessment
🇬🇧
FCA SYSC 6.3 + JMLSG
UK Risk-Based Approach
🇺🇸
FinCEN CDD Rule
US Customer Risk Profile
🇪🇺
EU 6AMLD Art 8
EU Risk Assessment Standard
🇭🇰
HKMA AMLO Sch 2
HK Risk-Based Methodology
🇦🇺
AUSTRAC ML/TF Risk
AU Programme Standards
🇦🇪
DFSA Module AML §7
UAE Risk Assessment Rules
Part of the Platform

Risk Assessment, Wired Into Every Customer Decision

The customer risk score is the central input to the rest of the platform. It drives EDD triggering, transaction-monitoring thresholds, periodic review cadence, and the level of approval required for any material customer event.

🧭
Drives
EDD · TM · Periodic Review

A customer's risk classification doesn't just sit on the file — it controls everything downstream. High-risk customers automatically receive EDD requirements at onboarding, tighter transaction-monitoring rules, shorter review cycles and senior-level approval requirements for material changes.

The result is a risk-based approach that actually behaves like one — visible end-to-end, calibrated to each customer's profile, and defensible at the factor level to any supervisor.

Customer Risk Assessment FAQ

What Compliance Teams Ask Us

How is the risk matrix configured?+
The default matrix covers five dimensions (customer, product, geography, channel, behaviour) with configurable factor weighting per dimension. Customer books with specific risk profiles — for example, private banking with HNWI exposure, fund administration with institutional investors, or payments with corporate clients — typically customise the matrix calibration without changing the underlying methodology.
How often does the risk score recalculate?+
Scores recalculate on any material event: new transaction monitoring alert, jurisdictional risk update, sanctions or PEP status change, beneficial ownership change, periodic review completion. There is no fixed recalculation schedule — the score reflects current state.
Can we see why a customer is scored a particular way?+
Yes. Every risk score includes the factor-level breakdown — which dimensions contributed how much, which underlying values triggered each factor, and when each component last updated. Analysts and auditors see the calculation, not a black-box result.
How does the platform handle country / geographic risk?+
Country risk is layered from FATF lists (grey, black, and high-risk monitored jurisdictions), Basel AML Index scoring, Transparency International CPI, EU high-risk third country lists, and internal jurisdictional risk classifications. The combined country risk score updates automatically as inputs change.
Can we override an algorithmic risk score with analyst judgement?+
Yes. Analyst overrides are supported with mandatory justification and approval routing — typically requiring senior compliance or MLRO sign-off. Every override is logged with the analyst identity, reasoning and approval trail in the customer record.
What does the audit evidence look like?+
For every customer, the audit trail shows: the current risk score, the factor-level breakdown, the history of every score change with the triggering event and date, any analyst overrides with approval evidence, and the matrix version applied. Exportable in formats accepted by MAS, FCA, FinCEN and the other major regulators.

Risk Scores That Survive Examination.

See how One Constellation handles Customer Risk Assessment as a transparent, multi-factor, dynamically recalculated score — defensible to any regulator.

Scroll to Top