AI KnowledgeAI Knowledge helps keep your Comm100, Confluence, or ServiceNow knowledge base up to date by running two types of AI scans:
- Static Scan – It’s like copying an article into ChatGPT and saying, “Rewrite this to make it clearer!” The AI reviews selected articles and provides recommendations for:
- Grammar and typos
- Clarity
- Redundancy
- And more
- Feedback Scan – It analyzes unanswered or down-voted questions coming from AI Agents or AI Copilots. If there's a related article, it suggests specific revisions to improve that content. If not, it creates a ready-to-publish draft based on the question.
How to run an analysis
By clicking the “New Analysis” button, you can choose between two options:
- Static Analysis
Run an analysis on selected Providers (Comm100, Confluence, ServiceNow), specific Knowledge Bases, and chosen categories. - Feedback Analysis
Use this to analyze feedback from AI Agents or AI Copilots. To run an analysis you should decide about:- Source: Unanswered questions or thumbs-down responses from AI Agents or AI Copilots
- Created Date: Filter by when the sources were generated
- Is Processed: Avoid re-analyzing the same sources by filtering out already processed items
How to check the recommendations (result of the analysis)
After waiting for the analysis to finish (It might take couple of minutes based on the number of articles, number of unanswered questions & thumb down answers) the status of the analysis changes to “Completed” from “In Progress”; now you can:
- Click on the number of recommendations in the Analyses page
- Filter the related recommendations under “Recommendations” page.
How to Use Recommendations
Each analysis generates a number of recommendations (ranging from none to hundreds). There are two types:
- Revision: Suggestions for improving an existing article.
Note: Static analysis results only include revision recommendations.
- New Article: Created when the AI can’t find a related article. It generates a draft for a brand-new article based on the identified need.
If the AI needs more information before drafting the article, it will first ask some clarifying questions. The agent can then decide whether to proceed with generating the article based on those answers.
Note: In this version, the original article is not updated automatically. Agents can review the recommended content and manually apply the changes to the original article as needed.
Permissions
Name | Description | Default of Agent |
View AI Knowledge Analyses | View AI Knowledge analyses | Yes |
Manage AI Knowledge Analyses | Create/Delete AI Knowledge analyses | No |
View and Manage AI Knowledge Recommendations | View AI Knowledge recommendations Edit/Delete AI Knowledge recommendations | Yes |
AI Quality Assurance
The AI QA & Coaching feature reviews each customer chats and gives it a score based on selected knowledge base. It shows what went well, what didn’t, and why—using plain explanations instead of technical jargon. Managers can quickly turn any issue into a coaching note for the right agents, and agents can see their own feedback to improve next time.
How to run a “Review”
After clicking the “New Review” button, you can filter the chats you want the AI to review based on the following criteria:
- Time Range
- Departments
- Agents
- Knowledge Domain
- Provider (e.g., Comm100, Confluence, ServiceNow)
- Knowledge Base (KB)
How to Check the “Review Items” (Results of Reviews)
After a few minutes (or possibly hours, depending on the number of chats), the review status will change from “In Progress” to “Finished.”
There are two ways to view the results of a review (also known as Review Items):
- Click on the number under the “Review Items” column for the specific review.
- Use the “Review” filter in the Review Items tab to filter and view results related to a specific review.
How to use “Review Items”
By clicking on “Details” under the operations column u can check details of each “Review Item”
How to create a “Coaching Opportunity”
Managers can highlight specific review items as learning moments for their team. To do this:
- Click on “Create Coaching Opportunity” on any review item.
- Assign the opportunity to one or more agents.
- (Optional) Add comments or tags to provide additional context or categorize the opportunity.
How agents can check assigned “Coaching Opportunities”
Each agent can see their assigned opportunities under the “Coaching” tab > “Assigned Opportunities”
How managers can track different “Coaching Opportunities”Under the Opportunity Management tab, managers can track all coaching opportunities and monitor their Completion Rate.
To view assigned opportunities, click on the blue “Assignees” filter—this will show all opportunities assigned to each agent.
Permissions:
View AI Quality Assurance Reviews | View AI quality Assurance reviews | No |
Manager AI Quality Assurance Reviews | Create/Delete AI quality assurance reviews | No |
View All Assigned Opportunities | View all coaching opportunities assigned to the agents of the site | No |
