Expand my Community achievements bar.

Submissions are now open for the 2026 Adobe Experience Maker Awards.

AI Assistant for Workfront

Avatar

Level 1

I would like to suggest a few items regarding AI Assistant for Workfront: 1) can you provide videos and sample prompts to accompany the AI Assistant for Workfront documentation? The documentation is too generic and 2) will you have a best practices session on AI Assistant for Workfront?  Especially around prompts 

1 Reply

Avatar

Level 5

Hi Liz - please review the Skill Exchange session that Kari Woolf and I did a few weeks ago regarding this topic.  I believe the videos were just posted on Experience League.

 

But to help you along.   The most important thing to remember is creating the prompt.  In the presentation I gave 4 simple steps in creating a prompt:

 

1) Start with what you want [formula, health, summarize, etc];

2) Provide some details on your want [identify two fields, project name, object, etc];

3) Expected results of your request [show number of days, display results, etc] 

4) Rephrase or ask additional questions [refine the list, be more specific on, etc]

 

Prompting is an art form - sure there are some technical aspects to a prompt.   Here are a few prompt examples:

 

Create a formula that shows the difference in the planned complete date and the actual complete date.

Summarize my upcoming work.

Generate textmode to pull the USER NAME into an issues report.

List all projects that contain "Red Bull" in the project description. 

Show me Project Health for this project.

 

This article can also help with keywords to use in your prompt:

Key Words for Prompting 

 

Depending on your expected result, you may need to be on the object which holds the information you are looking for, such as a Portfolio, Program, Project, Task, Issue, Request, etc. when using the AI Assistant.

 

It is a good idea to start a catalog of successful prompts for your company.   Sharing your prompts is a great way understand what you are asking the AI Assistant and the expected outcome to help with the User Experience. 

 

Lastly, don't forget to praise the AI Assistant (thumbs up) when it shows you the expect result or tell the AI Assistant it was wrong (thumbs down) along with feedback on your expected result.   This is how the AI Engineers can help the AI Assistant become more accurate and useful.

 

I hope this helps.