Decision Engine
The Decision Engine in Vantage enables automated decision-making within workflows through conditional logic, compliance checking, and AI-powered evaluation.
Conditional Routing
Multi-Conditional Node
The Multi-Conditional node evaluates data against multiple conditions and routes rows to different downstream paths:
- Define conditions using field values, operators (equals, greater than, contains, etc.), and thresholds
- Each condition has its own output port, allowing different processing paths
- Rows that don't match any condition can be routed to a default "else" path
Example: Route customer orders based on value — orders over $1,000 go to "High Value" processing, orders under $100 go to "Small Order" processing, everything else goes to "Standard."
Logical Trigger
The Logical Trigger starts a workflow when specific conditions are met:
- Evaluates field values against defined criteria
- Supports AND/OR logic for combining conditions
- Acts as a gateway that only fires when all conditions are satisfied
Compliance Checking
Compliance Check Node
The Compliance Check node evaluates data rows against predefined compliance rules:
- Define rules with field-level criteria
- Each row is evaluated and marked as PASS or FAIL
- Two output ports: one for passing rows, one for failing rows
- Useful for data quality checks, regulatory validation, and policy enforcement
AI Compliance Check Node
The AI Compliance Check node uses AI to evaluate compliance against natural-language policies:
- Define your compliance policy in plain English (e.g., "All records must have a valid email and consent date within the last 12 months")
- AI evaluates each row against the policy
- Returns PASS/FAIL results with explanations
- Supports batch processing for large datasets
Example Policy:
Check that:
1. Customer email is in a valid format
2. Consent date is not older than 12 months
3. No personally identifiable information appears in the notes field
AI-Powered Decisions
AI Conditional Node
The AI Conditional node uses AI to make routing decisions based on natural-language criteria:
- Describe the condition in plain English
- AI evaluates each data row and routes it accordingly
- Supports true/false output ports for binary decisions
- More flexible than rule-based conditionals for ambiguous or complex criteria
Example: "Route this customer inquiry to the billing team if it's about payments, invoices, or subscription issues. Otherwise, route to general support."
AI Trigger Node
The AI Trigger starts a workflow based on AI-detected patterns or events:
- Define what the AI should watch for
- AI evaluates incoming data and triggers the workflow when the condition is met
- Useful for anomaly detection, sentiment shifts, and pattern recognition
Building a Decision Pipeline
A typical decision pipeline combines multiple decision nodes:
Data Source → Data Validation → AI Compliance Check → Multi-Conditional
│ │
▼ ▼
FAIL: Send Alert PASS: Dashboard Output
- Data Source — Pull data from a database or integration
- Data Validation — Check data format and completeness
- AI Compliance Check — Evaluate against compliance policies
- Multi-Conditional — Route results based on outcomes
- Output — Alert on failures, display passes on dashboard