Solution Overview
AI Assist for service desk
By integrating AI Assist into service desk workflows, organizations can significantly improve ticket management, accelerate resolution times, and enhance customer support. Key capabilities include:
- Classification, which generates automated category, impact, and urgency proposals for faster ticket management.
- User Mood, which uses sentiment analysis to facilitate tailored responses.
- Summary, which provides ticket summarization for streamlined processing.
- Resolution Helper, which recommends knowledge base articles as well as similar solved tickets to help agents solve issues faster.
All of these features can be accessed via the AI Assist toolbar, which is located in the lower right corner of the screen in the Tickets, Incidents, and Service Requests sections.
Classification
As service desk agents review tickets, the system suggests the appropriate category, impact, and urgency for each ticket. Agents can review these suggestions and apply or dismiss them independently. This functionality speeds up ticket management by providing pre-calculated priorities. These recommendations are made only when no category, impact, or urgency has been set and are not re-evaluated after being accepted or dismissed.
The classification of the ticket impacts three properties of the ticket:
- Category
- Urgency
- Impact
Whenever a ticket is opened and the Classification tab is selected in the AI Assist toolbar, values are suggested for each of the three properties mentioned above. These suggestions are based on the ticket's contents and can be collectively accepted or dismissed individually.
User Mood
This feature helps agents understand users' emotional states, enhancing the quality of communication. The system analyzes the ticket’s subject and description and adjusts sentiment insights as new communications are added. This allows agents to modify their responses to better align with the user’s emotional state, improving the overall experience.
Three possible sentiments are identified:
- Positive - The customer appears satisfied. Maintain a positive tone and reinforce good service.
- Neutral - The customer's tone appears balanced. Address any concerns promptly to ensure a positive outcome.
- Negative - The customer appears dissatisfied. Careful communication is advised.
The sentiment for the ticket is updated whenever there is a new communication between the agent and the end user in the Journal section of the ticket. You can at all times see the current sentiment, as well as the previous one, helping you see the way the ticket has evolved.
The update is performed only upon reopening the ticket in the Service Desk and selecting the User Mood section in the AI Assist toolbar.
Summary
This feature enables service desk agents to generate structured summaries of tickets, providing a clear overview to enhance efficiency. The summary includes:
- Overview: based on subject, description, responsible agent, category, priority, and affected users.
- Communication: based on inbound and outbound interactions.
- Actions taken: based on ticket forwards, pauses, etc.
When generating the summary, only the last 5 actions on a ticket are considered.
The summary is stored for later use and can be easily copied for reporting purposes. The summary can also be regenerated by selecting Regenerate Summary in the AI Assist toolbar.
Resolution Helper
This feature helps support agents solve tickets faster by using readily-available knowledge and similar tickets matching the current ticket. To do this, based on the content of the ticket, the AI suggests knowledge base articles, as well as similar tickets that are already solved, giving quick solutions to the issues at hand:
Knowledge base articles - The knowledge base is searched for articles that are relevant to the issue. These articles can be easily accessed from the AI Assist toolbar, making the use of knowledge easier.
Similar closed tickets - The tickets that are similar to the one that is viewed by the Agent are recommended in the AI Assist toolbar, making it easier to find already-existing solutions for the issue at hand.
Only 3 Knowledge base articles and similar closed tickets are displayed in the Resolution Helper.