- Post History
- Subscribe to RSS Feed
- Mark as New
- Mark as Read
- Bookmark
- Subscribe
- Printer Friendly Page
- Report Inappropriate Content
22m ago
Instantly create summaries, actions, and follow-ups in the activity stream
Activity Response Generation is a generative AI capability in Now Assist for CSM that enhances the activity stream experience by allowing human agents to instantly generate high-quality, context-aware updates directly within Comments and Work Notes. Instead of writing repetitive updates manually, human agents can generate clear, structured messages grounded in case data, improving communication speed and consistency while reducing effort.
This capability uses the content of the case record, including Short description, Description, Additional comments, Work notes, State, Priority, and any included emails to produce accurate and relevant output.
Key Capabilities
-
Context Menu actions for common updates: Human agents can trigger out-of-the-box actions directly from the activity stream: Summarize actions taken, Follow up, Post response, and Acknowledge task.
-
Full case context awareness: The skill interprets the entire activity stream and case context, including optionally emails, to draft updates that reflect the current situation with precision.
-
Agent control over final output: Generated content serves as a starting point. Human agents review, edit, and approve all messages before posting.
-
Customizable through Skill Kit: Administrators can tailor tone, style, and required elements to align with organizational communication standards.
Usage walkthrough:
Implementation
-
Find the skill in the Customer workflow: Open Now Assist Admin Console and go to the Customer workflow.
-
Configure the skill in the Guided Setup:
-
Choose inputs: Confirm Case as the input table. Choose the input fields: Short description, Description, Additional comments, Work notes, State, Priority. Enable the option to include emails so email content is also used as grounding.
-
Define access: Specify which roles and teams can see this NACM action. Optionally restrict access by channel or case condition so it appears only when it adds value.
-
Select display: Allow the Now Assist Context Menu actions to show up in product for the specific roles you selected on the previous step
-
Review and activate: Confirm configuration, then activate so it becomes available for agents.
-
-
Fine-tune in Now Assist Experiences and Skill Kit (optional)
-
In Now Assist Experiences, refine additional NACM options such as availability rules, options, and placement.
-
In Now Assist Skill Kit, clone the skill and prompt per provider, add conditions or tools, and finalize and test the new configuration. Example here.
-
-
Validate and roll out: Test the action on sample cases to confirm quality of generated content. Roll out to selected teams first, gather feedback, then expand usage to more human agents.
Configuration walkthrough:
Key Best Practices
-
Keep work notes and comments clear and complete to improve grounding accuracy
-
Encourage human agents to use generated content as a baseline and adjust as needed for specific customer situations
-
Use Skill Kit to customize tone and phrasing to match your organization's voice and continue to iterate as needed
-
Apply role-based availability so only relevant personas see this capability
-
Review human agent feedback regularly to refine phrasing or improve field completeness
-
Start with a pilot team to gather feedback before broader rollout, use Now Assist Analytics to improve usage
Formal Learning
- Now Learning: Now Assist for CSM Essentials Path
- AI Essentials
- Introduction to Generative AI
- Now Assist Essentials
Measured Success & Outcomes
| Outcome | How It Supports Agents | Key Metric |
|---|---|---|
|
Reduced manual writing |
AI drafts comments and work notes quickly, removing repetitive effort |
Time saved per update |
|
Improved communication consistency |
Standardized structure and tone across activity entries |
Reduction in messaging variation |
|
Faster case progression |
Less time spent writing and formatting updates |
Time between activity log entries |
|
Better human agent efficiency |
Removes repetitive tasks and boosts productivity |
Human agent satisfaction and adoption rate |
|
Increased collaboration |
Clear, contextual communication visible to all team members |
Quality of internal handoffs |
Frequently Asked Questions
What content does the skill use to generate responses?
It analyzes Short description, Description, Additional comments, Work notes, State, Priority, emails, and the entire activity stream to produce context-aware output.
Can admins customize the generated output?
Yes. You can tailor tone, style, and required elements using Skill Kit. Clone the skill and prompt to create configurations specific to your organization's needs.
Does this replace the need for agents to review messages?
No. Human agents always review, edit, and approve the generated output before posting. The AI provides a draft, not a final message.
Which use cases benefit most?
High-volume, repetitive updates such as follow-ups, action summaries, cross-team routing, resolution acknowledgments, and task status updates see the greatest efficiency gains.
How is this different from case summarization?
Case summarization creates a summary of the entire case for human agent understanding. Activity Response Generation creates individual activity entries (comments or work notes) for ongoing case communication.