Business Content in CX Enterprise Coworker (CitiSignal)
This demo is sourced from the Adobe Tech Insider hands-on exercise.
You're encouraged to use those training materials to learn more about Adobe Agentic AI capabilities.
ο»Ώ
Demo Script: Business Context in CX Enterprise Coworker
Scenario: Grounding the Coworker agent in CitiSignal business knowledge, then using it to answer business questions and build/publish an audience.
Demo Environment:
- IMS Org: Adobe Demo System Shared org
- AEP Sandbox: Public Citi-signal (Telco)
Est. time: 8β12 min
Feature stage: π‘ Beta
ο»Ώ
Step 1: Pre-flight checklist
- Logged in to CX Enterprise Coworker under the Adobe Demo System Shared org.
- Public CitiSignal (Telco) sandbox selected.
- Coworker opens, and the Business Context item is visible in the left menu.
- Decide your audience naming token β the lab uses your LDAP
ο»Ώ
Step 2: Verify the context with live prompts
Open + New chat and run these in order. Each prompt proves the agent is drawing on the uploaded document, not general knowledge.
ο»Ώ
# | Prompt | What to point out |
|---|---|---|
1 | what are the CitiSignal personas? | Personas come straight from the uploaded doc β proof the context loaded. |
2 | when is CitiSignal's Tax Refund season? | A business-specific fact the model couldn't know otherwise. |
3 | How did my website activity change during tax season this year? | Agent now combines business context with the sandbox's behavioral data. |
4 | Create an audience of customers who have high disposable intent in the last 30 days. | "High disposable intent" is defined in the business doc β the agent applies your definition. |
Then confirm the build: 5. Type yes 6. Type yes (second confirmation)
7. Rename and publish β swap in your LDAP:
change the name to LDAP - High Disposable Intent - Last 30 Days and publish the audience
β Result: The audience LDAP - High Disposable Intent - Last 30 Days is created and published.
ο»Ώ