Skip to main content

Start Event Node

The entry point for every workflow - configure triggers, execution mode, and initial data. The Start Event Node is the required entry point for every workflow. It defines how the workflow is triggered, how it executes, and what initial data it receives.
Every workflow must have exactly one Start Event Node. It’s automatically added when you create a new workflow.

Overview

The Start Event Node controls:

Configuration

Click on the Start Event Node to open its configuration panel.

Execution Mode

Toggle Start asynchronously to control how your workflow responds when triggered:
Synchronous workflows can timeout if processing takes too long. Use async for any workflow that calls external APIs or performs complex operations.

Default Context

The context is your workflow’s data storage—a JSON object that travels through every node. Configure default values in the Start Event Node. Click + Set Default workflow content to define the initial structure:

System Properties

Every workflow execution automatically includes these system properties: Example context at workflow start:
Use correlationId to trace related workflows. When one workflow triggers another, they share the same correlation ID—useful for debugging complex processes.

Manual Triggering

The Start Event Node displays cURL commands for manually triggering the workflow via HTTP. This is useful for testing outside of Preview mode. QA Environment:
Production Environment:
The {slug} is the unique workflow identifier visible in your browser URL when editing the workflow.

Styling

Customize the node’s visual appearance on the canvas: Styling is optional and doesn’t affect workflow execution—it’s purely for visual organization when building complex workflows.

How Data Enters the Workflow

Data can enter through the Start Event Node in several ways:

From Voice AI

When Voice AI triggers a workflow (via a Tool), it passes the conversation transcript and any extracted data:

From API Request

When triggered via HTTP API, the request body becomes part of the context:
Results in context:

From Another Workflow

When one workflow calls another, it can pass data through the invocation:

Best Practices

Define Clear Default Context

Set up a default context structure that matches what your workflow expects:
This makes it clear what data the workflow needs and provides a consistent structure for downstream nodes.

Use Async for External Calls

If your workflow calls any external API (Snapsheet, Guidewire, your AMS), use asynchronous execution to avoid timeouts.

Document with Styling

Use colors to categorize workflows visually:

Common Issues

Workflow Times Out

Symptom: Synchronous workflow returns a timeout error. Fix: Switch to asynchronous execution, or optimize the workflow to complete faster.

Missing Context Data

Symptom: Downstream nodes can’t find expected data. Fix: Verify that the triggering system (Voice AI, API) is sending the expected payload. Check the default context structure.

Wrong Environment

Symptom: Workflow connects to QA systems when it should use Production. Fix: Check the environment property in context. Ensure you’re triggering the correct environment endpoint.

Next Steps

Context Deep Dive

Learn how data flows through workflows

Mapping Node

Transform data after it enters the workflow

Task Node

Call external systems from your workflow

End Event Node

Complete the workflow and return responses