Skip to main content

Workflow Nodes

Nodes are the fundamental building blocks of a Datalex workflow. Each node represents a discrete step in your fraud detection or routing logic. Nodes are connected by edges to form a directed graph that is evaluated in real time against incoming transaction data. Workflow Canvas Overview The Datalex workflow canvas with a sample node graph.

Node Categories

The node palette is organized into three categories: Triggers, Integrations, and Actions. Drag any node from the side panel onto the canvas to add it to your workflow.
CategoryPurpose
TriggersEvaluate conditions and route execution
IntegrationsEnrich transactions with external or computed signals
ActionsTerminate a branch with a final decision or side-effect

Input Node

Every workflow begins with exactly one Input Node. This node defines the transaction attributes that are available to all downstream nodes. Input Node The Input Node with configured attributes. How it works:
  • Attributes are declared as typed fields (e.g., Text, Number, Boolean).
  • Any attribute added here becomes available in condition rules throughout the workflow.
  • Common attributes include payment details, merchant information, card data, and user context.
Configuration:
  1. Click the Input Node on the canvas to open the Attribute Editor.
  2. Add attributes by specifying a name, type, and description.
  3. Elements added here are added as additional params to eb used in input fields on top of the predefined ones
  4. Save — the attributes are immediately available to all Condition Nodes.

Condition Node

A Condition Node evaluates one or more rules against the transaction data and routes execution to different branches based on the outcome. Condition Node A Condition Node with two rules configured: a passing branch (#1) and a failing branch (#2). Rule structure: Each rule consists of:
  • Attribute — the data field to evaluate (e.g., paymentInfo.amount)
  • Operator — the comparison logic (see table below)
  • Value — the threshold or target value
Supported operators:
OperatorDescription
EQUALExact match
NOT_EQUALDoes not match
GREATER_THANNumeric greater-than comparison
LESS_THANNumeric less-than comparison
CONTAINSString contains substring
STARTS_WITHString starts with prefix
ENDS_WITHString ends with suffix
IS_NULLField is absent or null
IS_NOT_NULLField is present and not null
Outputs:
  • Branch #1 (green) — all rules evaluate to true.
  • Branch #2 (red) — one or more rules evaluate to false.
  • Additional custom branches can be added for more complex routing.
A Condition Node must have at least one fully configured rule before you can save the workflow.

Action Nodes

Action Nodes terminate a branch with a decision or trigger a side-effect. They have no outgoing connections.

Allow

Marks the transaction as approved. Execution stops on this branch. Allow Node

Reject

Marks the transaction as rejected. Execution stops on this branch. Reject Node

Node Options

Selecting a node allows for specifying additional parameters such as flagging the transaction or adding an insight which is like a note. Additionally, a toggle option is provided to create a case. Node Options

Create Case

Opens a new investigation case in the Case Management module and links it to the triggering transaction. Case Options Configuration options:
  • Case name / reason — a descriptive label for the generated case.
  • PriorityLOW, MEDIUM, HIGH, or CRITICAL.
  • Assigned investigator — optional pre-assignment.

Integration Nodes

Integration Nodes enrich the transaction context with computed or externally fetched signals. Their outputs become available as attributes in downstream Condition Nodes.

Velocity

Counts how many times a given attribute value (e.g., a card fingerprint) has appeared within a configurable rolling time window. Velocity Node Configuration:
  • Attribute — the field to track (e.g., card.fingerprint).
  • Time window — the look-back duration (e.g., 5 minutes, 1 hour, 24 hours).
  • Threshold — the count that defines high velocity.
Output: The computed velocity count is injected as a new attribute that can be referenced in subsequent Condition Nodes.

Device

Performs device fingerprint analysis to detect anomalous device behaviour, reused devices across multiple accounts, or unusual browser/OS combinations. Device Node Evaluated signals:
  • Device fingerprint
  • Browser and operating system
  • Device velocity (number of transactions from the same device in a time window)
  • Known suspicious device flags

IP Lookup (VPN Detection)

Resolves the transaction IP address and checks whether it originates from a known VPN, Tor exit node, data-centre range, or IP blocklist. IP Lookup Node Configuration:
  • Toggle individual detection categories (VPN, Tor, proxy, data-centre).
  • Set the desired action on a positive match (route to specific branch).
Output attributes:
  • vpn.isVpn — boolean
  • vpn.isTor — boolean
  • vpn.isProxy — boolean
  • vpn.country — detected country from IP geolocation

Country Block

Allows or restricts transaction processing based on the geographic origin of the transaction. Country Block Node Configuration:
  • Select a list of allowed or blocked countries.
  • Choose whether unrecognized countries are treated as blocked or allowed.

Custom API (Integration)

Calls an external REST API and maps the response fields to workflow attributes, enabling integration with any third-party data source. Custom API Node Configuration:
  • Endpoint URL — the target REST endpoint.
  • HTTP methodGET or POST.
  • Headers and payload — static or dynamic values using {{ attribute }} syntax.
  • Response mapping — define which response fields become workflow attributes.

Sub-Workflow

Embeds a previously saved workflow as a reusable step inside the current workflow. Sub-Workflow Node Use cases:
  • Shared compliance checks reused across multiple workflows.
  • Modular fraud checks that can be updated independently.
Configuration:
  • Select the target workflow from the dropdown list.
  • The sub-workflow receives the same input attributes as the parent.

Node Validation Rules

Before a workflow can be saved or activated, the following structural rules are enforced:
  1. Exactly one Input Node must be present.
  2. At least one Action Node (Allow, Reject, etc.) must be reachable.
  3. Every node except the Input Node must be connected to at least one edge.
  4. All Condition Nodes must have at least one fully configured rule (attribute, operator, and value).
  5. Sub-Workflow Nodes must reference a valid, saved workflow.
Validation errors are surfaced inline on the canvas and in the save dialog.

Adding a Node

  1. Open the node palette from the left-hand sidebar.
  2. Drag the desired node type onto the canvas.
  3. Connect it to an existing node by dragging from the source handle (bottom of a node) to the target handle (top of a node).
  4. Click the node to open its configuration panel.
  5. Save the workflow.
Adding a Node Dragging a Condition Node from the palette onto the canvas.