RiskOS™ Dashboard Setup

Learn how to set up Graph Intelligence in the RiskOS™ Dashboard to detect fraud signals and optimize decisions.

Set up Graph Intelligence in the RiskOS™ Dashboard

Before you start

Graph Intelligence is available as an add-on to select RiskOS™ enrichments. To use it, add one of the following enrichments to your workflow with Graph Intelligence enabled:

  • Sigma Identity Fraud + Graph Intelligence
  • Verify + Graph Intelligence
  • Email Risk + Graph Intelligence
  • Phone Risk + Graph Intelligence

Availability depends on your RiskOS™ configuration and enabled products.


How it works

Graph Intelligence provides additional signals that reflect how identity attributes relate to other entities within a broader network context.

These signals can help surface patterns such as reuse, connectivity, or unusual relationships that may indicate elevated fraud risk.

Graph Intelligence outputs are returned alongside the primary enrichment and can be used to inform workflow routing and review decisions.


Graph Intelligence outputs

Graph Intelligence returns structured features associated with identity attributes evaluated in the request.

  • Features are grouped by attribute type (for example, identifiers or contact information).
  • Each group may include multiple signals describing observed relationships or activity patterns.
  • Most features are numeric and can be incorporated into workflow logic as part of a broader risk strategy.

Not all features are available for every evaluation. Availability depends on the inputs provided and the context in which they are evaluated.


Using Graph Intelligence in workflows

You can reference Graph Intelligence outputs when defining Conditions in your workflow to support routing decisions such as review, escalation, or additional verification.

Graph Intelligence signals are intended to complement other enrichment outputs and should be interpreted in context rather than used in isolation.


Considerations

  • Graph Intelligence signals are only available for attributes included in the request and successfully evaluated.
  • Some signals may be unavailable or incomplete when limited historical context exists for a given attribute.
  • Results may vary over time as new information becomes available.