Skip to main content
Build workflows that turn real-time computer vision detections into automated actions — events, alerts, emails, and more.

​
How it works

Every workflow follows the same three-step pattern:
1

Inject detection data

A Trigger Node receives real-time track and zone data from the Worlds detection pipeline via the state machine.
2

Apply business logic

Check Nodes evaluate conditions — is the object in a specific zone? Has it been there long enough? Is it moving too slowly? Chain checks together to build precise rules.
3

Take action

Event and Action Nodes create events in the Worlds UI, capture images, generate GIFs, send email alerts, and more.

​
Get started

Architecture Overview

Understand how detections flow from cameras through the state machine into your workflows.

Prerequisites

What you need before building your first workflow.

Node Reference

Detailed reference for every Worlds custom node.

Streaming Track State Guide

Build a real-time workflow that reacts to track updates.