ThetaRay
AI-led AML intelligence
ThetaRay focuses on AML monitoring, customer risk assessment, and sanctions screening. Older material refers to SONAR; the current site introduces the RAY Platform as an agentic financial crime intelligence layer. Treat product naming and migration scope as explicit diligence items.
Our assessment
Specialist AML intelligence and screening; do not assume full payment fraud, identity proofing, or underwriting.
Best fit and limitations
Worth evaluating for cross-border payment networks, correspondent banking, and payment providers with complex transaction flows. Compare its incremental discovery against your established scenario library.
- Anomaly detection can produce unusual but legitimate activity; measure investigator usefulness as well as novelty.
- Risk assessment uses KYC inputs and does not replace collecting and verifying identity.
- Confirm current product names, case features, report submission, and agent scope before procurement.
Tools 5
AML transaction monitoringBehavioral detection+
- AI-based analysis of money movements
- Detection intended to surface risks beyond fixed scenarios
Customer and transaction screeningSanctions controls+
- Check people, companies, and counterparties
- Sanctions, PEP, and adverse media screening
Customer Risk AssessmentCDD risk+
- Combine KYC attributes and transaction behavior
- Accept monitoring data from ThetaRay or another system
Screening risk controlsOperational configuration+
- Adjust matching rules and parameters
- Risk scoring and alerts for review
RAY investigation suitePost-alert investigation and analyst assistance+
- Combine case, KYC, transaction, counterparty, and historical context
- Prepare explained investigation findings and recommended next steps
- Support analyst questions and document summaries
AI capabilities
Anomaly detection + RAY investigation assistance
RAY gathers alert and customer data, compares behavior with KYC profiles, and prepares an investigation report. Its assistant summarizes documents and explains case context. Analysts retain control of final decisions.
What to validate
Test the source of each finding, missing-data handling, and analyst approval. Validate anomaly detection separately from the quality of generated reports.
Implementation
Integration checklist
- Supply sufficient transaction history and stable sender/receiver identifiers.
- Compare standalone monitoring with an overlay to the existing AML system.
- Test country, currency, corridor, and peer-group effects on risk results.
Commercial scope
Quote required. Clarify monitoring volume, history loads, screening, customer risk, cases, and RAY/agent entitlements.
Questions for the demo
- What does the model find that our current scenarios miss?
- Can every alert show the baseline and reason the activity is unusual?
- Which RAY functions are available now, and how do SONAR customers migrate?
Engineering
- Inputs
- Person and organization names with secondary identifiers · Transaction counterparty details · Existing KYC profiles, alert history, transactions, and adverse-media context through secure APIs · Behavioral transaction-monitoring alerts and their contributing evidence
- Outputs
- Candidate matching scores and screening alerts · Natural-language case file with retrieved evidence and an interactive analyst assistant · Findings, contradictions, KYC consistency checks, historical context, and next-step recommendations
- Webhooks
- Not verified in public documentation.
- Decision timing
- The screening product documents a REST API and real-time alerts. No contracted latency or measured throughput was verified.
- Deployment
- Microsoft Azure infrastructure; overlay connects to existing on-premises or cloud case/monitoring/screening systems
- Dependencies
- Source-system access and institution-owned final decisions · ThetaRay Investigation Center and enabled RAY features
Verified details 12
- Inputs
Person and organization names with secondary identifiers
Source Checked 2026-09-18 - Inputs
Transaction counterparty details
Source Checked 2026-09-18 - Outputs
Candidate matching scores and screening alerts
Source Checked 2026-09-18 - Decision timing
The screening product documents a REST API and real-time alerts. No contracted latency or measured throughput was verified.
Source Checked 2026-09-18 - Inputs
Existing KYC profiles, alert history, transactions, and adverse-media context through secure APIs
Source Checked 2026-09-18 - Outputs
Natural-language case file with retrieved evidence and an interactive analyst assistant
Source Checked 2026-09-18 - Deployment
Microsoft Azure infrastructure; overlay connects to existing on-premises or cloud case/monitoring/screening systems
Source Checked 2026-09-18 - Connected provider
Microsoft Azure: RAY infrastructure
Source Checked 2026-09-18 - Dependency
Source-system access and institution-owned final decisions
Source Checked 2026-09-18 - Inputs
Behavioral transaction-monitoring alerts and their contributing evidence
Source Checked 2026-09-18 - Outputs
Findings, contradictions, KYC consistency checks, historical context, and next-step recommendations
Source Checked 2026-09-18 - Dependency
ThetaRay Investigation Center and enabled RAY features
Source Checked 2026-09-18
Illustrative contract
An example for your internal adapter. This is not a vendor endpoint, request, or response.
{
"contract": "internal-risk-review/v1",
"vendorId": "thetaray",
"internalCorrelationId": "example-001",
"vendorRecordId": null,
"control": "Behavioral detection",
"evidence": [],
"status": "pending",
"decision": "review",
"requiresHumanReview": true
}Integration limits
- Illustrative internal integration contract. This is not a vendor request, response, endpoint, or SDK example.
- Empty country, deployment, or provider lists mean not verified in this review. A documented country refers to the specific product noted in its source, not universal platform coverage.
- Matching combines ML, deterministic rules, and secondary attributes. Match scores are not a tested probability of fraud. Authentication and callback guarantees are not published on this page.
- Support for on-premises source systems does not establish that RAY itself is available for on-premises hosting.
- Real-time detection language does not establish a synchronous API deadline for the full investigation.
Sources 7
Source review dates are shown above. Product claims come from public sources. Fit, limits, and evaluation questions are our analysis. This is not a hands-on performance test. Methodology · Changelog