AML / CTF Compliance

AML Shield Pro

A four-layer detection engine — Rules, AI Anomaly, Behavioural Analytics, and Network Intelligence — converging into a single Hybrid Risk Score with explainable factors and AUSTRAC-ready regulatory filing.

Start Free TrialSign In
Detection Engine

Four layers of intelligence. One explainable score.

Each layer independently evaluates the transaction stream, then results are weighted into a Hybrid Risk Score with plain-language reasons — not a black-box number.

01

Rules Engine

15+ typology rules covering structuring, smurfing, rapid fund movement, high-risk jurisdictions, and more.

Safety net — catches known typologies
02

AI Anomaly Detection

Unsupervised machine learning detects breakout behaviour and statistical outliers that rules cannot anticipate.

Catches what the rules miss
03

Behavioural Analytics

Dynamic customer baseline built from transaction history and peer group. Flags deviations in volume, velocity, and geography.

Personalised to each customer
04

Network Analytics

Entity relationship graph traces direct and indirect connections to sanctioned entities, criminal networks, and UBO structures.

Sees hidden relationships
Hybrid Risk Score
Rules (30) + Anomaly (20) + Behaviour (20) + Network (30) = 100-point scale. Threshold-configurable per entity type and business segment.
LOW
0–39
MEDIUM
40–69
HIGH
70–84
CRITICAL
85–100
Behavioural Intelligence

Not just “AI detected an anomaly.”

When Sentinel’s behavioural engine flags a customer, it shows exactly what changed and by how much — compared to their own history, their peer group, and statistical expectations.

That context is what makes an alert useful. It’s also what makes your SAR narrative defensible when AUSTRAC asks why the report was filed.

Behavioural Deviation — XYZ Trading Pty Ltd
Expected monthly volume
Expected $85,000 · Observed $412,000
+385%
Transaction frequency
Expected 18 / week · Observed 73 / week
+306%
Active jurisdictions
Expected 2 usual · Observed 7 detected
+5 new
New beneficiaries
Expected 1–2 / month · Observed 14 this week
+600%
BEHAVIOURAL ANOMALY — CRITICAL
Capabilities

Everything your AML programme requires

Real-Time & Batch Monitoring

Ingest from core banking APIs, payment feeds, ISO 20022 streams, or batch upload. Alerts generated within seconds of ingestion.

Detection Studio

Create, test, and simulate AML rules before deployment. See projected alert volumes and false-positive rates on your own historical data.

Entity 360 Investigation

Every alert opens an Entity 360 view: transaction history, behavioural profile, network graph, sanctions checks, adverse media, and case notes in one workspace.

Explainable Hybrid Risk Score

Four-layer score with plain-language contributing factors. Every score is traceable to the specific rule, anomaly, deviation, or network signal that drove it.

AUSTRAC-Ready Reporting

Auto-generate SMR and TTR reports in AUSTRAC-required format. Validate fields, record the approval workflow, and maintain e-filing submission audit trail.

Continuous Feedback Loop

Investigator decisions — true positive, false positive, SAR filed — feed back into rule thresholds and model calibration. Detection improves with every case closed.

Detection Studio

Build, simulate, and deploy AML rules — before they change your alert queue

Most AML systems require a professional services engagement to modify detection logic. Detection Studio lets your compliance team define rules, run simulations against historical transaction data, review projected alert volumes and false-positive rates — then deploy with version control and senior approval, no code required.

Before v4.2 of the Rapid Fund Movement rule went live, the team knew it would generate 2,381 alerts at an 11.4% false-positive rate. They adjusted thresholds until satisfied, then deployed with a single approval click.

Rule BuilderSimulation EngineVersion ControlApproval Workflow
Detection Studio · Rule v4.2
Rapid Fund Movement
APPROVED
Rule Conditions
Amount > $20,000
AND 3+ transactions within 24 hours
AND new beneficiary (first appearance)
Simulation Results · 1.2M transactions
Expected alerts
2,381
Est. FP rate
11.4%
Precision
88.6%
Version
v4.2
Deploy v4.2
Adjust Thresholds
Approved by: Compliance Officer · 14 Aug 2025 · Change #47
Compliance Operations

Monitor the programme, not just the alerts

Enterprise AML requires visibility into investigator throughput, SLA compliance, rule performance, and backlog health — not just individual case outcomes.

Alerts today
842
+12% vs 30-day avg
Open cases
127
18 high-risk priority
Avg investigation time
3.2d
SLA target: 5 days
High-risk backlog
18
Requires escalation
False-positive rate
14.7%
Down from 18.3% last month
SAR recommended
9
Pending compliance officer review
Rule Performance — Alert-to-Case Conversion Rate
Rapid Fund Movement v4.278% conversion · 312 alerts
Structuring — Sub-threshold Splitting91% conversion · 87 alerts
High-Risk Jurisdiction Transfer64% conversion · 204 alerts
Cash Intensive Business Anomaly83% conversion · 143 alerts
The Workflow

From ingestion to AUSTRAC filing

01
Ingest
Transaction data enters via API, batch feed, or ISO 20022 stream. Customer profiles and counterparty data are mapped to the entity graph.
02
Detect
All four detection layers run simultaneously. Results are weighted into the Hybrid Risk Score with contributing factor breakdown.
03
Alert
Scores above threshold generate structured alerts with contributing factors, behavioural deviations, and network connections surfaced for the analyst.
04
Investigate
Analyst reviews the Entity 360 view — transaction history, risk factors, network graph, sanctions checks, adverse media. Case notes recorded.
05
Report
SAR/SMR narrative auto-populated from case data. Validated against AUSTRAC requirements. Approval workflow logged. E-filing tracked.
06
Improve
Disposition feeds back into the detection engine. Rule thresholds self-calibrate. False-positive rate declines over time.

See the detection engine in action

Book a 30-minute demo and walk through a live alert — from transaction ingestion to AUSTRAC filing.

Request a DemoSign In