Transaction Layer

TxnMonitor-AI

Active

Periodic and on-demand transaction monitoring across a configurable lookback window. Surfaces AML patterns, fraud signals and velocity anomalies in a prioritised alert queue. Designed for scheduled overnight runs and periodic customer review — complementary to the real-time AML-AI plugin.

Endpoint

POST/api/txnmonitor/scan

Detection modules

AML Pattern Detection

Structuring, hawala, layering, rapid movement, cash-intensive business patterns detected across the full lookback window.

Fraud Signal Detection

Velocity anomalies, device switching, unusual merchant patterns, account behaviour changes.

Velocity Monitoring

Threshold breach detection — alerts when transaction count or total volume crosses jurisdiction reporting thresholds over the lookback period.

Request parameters

Parameter	Type	Required	Description
`customer_id`	string	Yes	Customer to scan. Your internal reference.
`jurisdiction`	string	Yes	One of: CA (Canada), AE (UAE), SA (KSA). Sets reporting thresholds and rules.
`lookback_days`	number	No	Days of history to analyse. One of: 7, 30, 60, 90 (default: 30).
`alert_threshold`	string	No	Minimum alert severity to surface. One of: low, medium, high, critical (default: medium).
`include_aml`	boolean	No	Enable AML pattern detection module (default: true).
`include_fraud`	boolean	No	Enable fraud signal detection module (default: true).
`include_velocity`	boolean	No	Enable velocity monitoring module (default: true).

Response fields

Parameter	Type	Required	Description
`success`	boolean	Yes	Whether the API call succeeded.
`data.transactions_scanned`	integer	Yes	Total number of transactions analysed in the lookback window.
`data.alerts`	array	Yes	Array of alerts. Each has severity, type, description, amount, and rule_id.
`data.alerts[].severity`	string	Yes	Alert severity: low, medium, high, or critical.
`data.alerts[].type`	string	Yes	Alert typology, e.g. structuring, velocity_breach, account_behaviour_change.
`data.alerts[].description`	string	Yes	Human-readable description of the alert.
`data.alerts[].amount`	number	No	Transaction amount associated with the alert, if applicable.
`data.alerts[].rule_id`	string	Yes	Rule identifier for the compliance audit trail.
`data.high_risk_count`	integer	Yes	Number of alerts at high or critical severity.
`data.ai_recommendation`	string	Yes	Prioritised next steps based on the alert queue.
`processing_time_ms`	integer	Yes	Total processing time in milliseconds.

AML vs TxnMonitor

AML-AI is called at the point of each transaction — real-time, sub-300ms.
TxnMonitor-AI is called periodically to scan transaction history — typically nightly or weekly — and surfaces patterns that only become visible across multiple transactions over time. Both should be used together for complete coverage.

Code example

curl -X POST \
  https://shield.kodexa.systems/api/txnmonitor/scan \
  -H "Content-Type: application/json" \
  -d '{
    "customer_id": "CUST-AE-001",
    "jurisdiction": "AE",
    "lookback_days": 30,
    "alert_threshold": "medium",
    "include_aml": true,
    "include_fraud": true,
    "include_velocity": true
  }'

Example response — alerts found

json

{
  "success": true,
  "data": {
    "transactions_scanned": 47,
    "alerts": [
      {
        "severity": "high",
        "type": "structuring",
        "rule_id": "TXNMON-STRUCT-001",
        "description": "4 transactions between AED 48,000–54,000 within 10 days, total AED 207,000. Pattern consistent with structuring to avoid AED 55,000 goAML threshold.",
        "amount": 207000
      },
      {
        "severity": "medium",
        "type": "velocity_breach",
        "rule_id": "TXNMON-VEL-003",
        "description": "Transaction frequency increased 340% vs prior 30-day baseline.",
        "amount": null
      }
    ],
    "high_risk_count": 1,
    "ai_recommendation": "1 high-severity alert requires immediate MLRO review. Structuring pattern should be assessed for goAML STR filing. Consider account restrictions pending review.",
    "request_id": "txnmon_1716000000_abc123",
    "created_at": "2026-05-22T04:00:00.000Z"
  },
  "processing_time_ms": 1203,
  "jurisdiction": "AE",
  "plugin": "txnmonitor"
}