| Audience | Platform admins and data teams, Marketers |
| Prerequisites | AI Decisioning overview → |
AI Decisioning (AID) helps you automate campaign decisions at the individual user level.
You create agents that replace traditional campaigns by combining audiences, outcomes, messages, and guardrails.
This guide covers:
-
Setup for platform admins and data teams →
Configure your workspace, prepare event data, and connect destinations. -
Day-to-day usage for marketers →
Build and manage agents, QA with Inspector, add Collections, and measure performance in Insights.
Setup for platform admins and data teams
Before marketers can build and launch agents, your workspace must be configured and connected to the right data and destinations. Most of this setup is done once per workspace.
Configure workspace and data
| Step | What you’ll do | Article |
|---|---|---|
| Set configuration | Define global eligibility rules, time zones, and channel defaults. | Configuration → |
| Define models | Define your parent audience, structure event models, and prepare tables for AID. | Prepare data for AID → |
Connect destinations
| Step | What you’ll do | Article |
|---|---|---|
| Set up destinations | Connect your messaging platform so AID can deliver actions. | Connect SFMC, Braze, or Iterable. |
How marketers use AI Decisioning
Once setup is complete, marketers can create agents to personalize campaigns without managing manual rules or journeys.
Build and launch agents
| Step | What you’ll do | Article |
|---|---|---|
| Create agents | Define campaign goals, target audiences, and message options. | Agents → |
| Add actions | Configure messages (emails, SMS, push notifications) to test and deliver. | Actions → |
QA and validate
| Step | What you’ll do | Article |
|---|---|---|
| Use Inspector | Spot-check how an agent will deliver messages for real users. | Inspector → |
| Add Collections | Dynamically recommend personalized products or content. | Collections → |
Measure and optimize
| Step | What you’ll do | Article |
|---|---|---|
| Review Insights | Analyze performance by user, message, and timing to refine agents. | Insights → |