Agent copilot is the name for a set of AI features that are designed to make agents as efficient as possible. By providing AI-powered context and suggestions, agent copilot helps agents solve customer requests more quickly and easily.
The following features are included in agent copilot:
Auto assist
Auto assist uses a large language model (LLM) to understand the contents of submitted tickets and make suggestions to your agents on how to solve them. It can also automatically perform actions that have been approved by an agent. With auto assist, agents spend less time on repetitive tickets, solve tickets in a more consistent way, and ultimately close more tickets.
Auto assist will tap into similar tickets that were solved successfully when procedures or public help center articles aren't available. Similar tickets will be used to suggest replies, allowing auto assist to continue suggesting helpful and accurate responses.
Auto assist can also be available for agents during phone calls. It will show AI suggestions (based on your help center content) in the Knowledge tab during phone calls.
To ensure auto assist is active on your tickets, create a trigger that assigns a tag to every ticket that is created. The tag should be 'agent_copilot_enabled'.
Suggested first replies
Suggested first replies use generative AI to suggest a first response for agents in tickets based on existing macros and help center articles. Using AI-generated suggestions to respond to tickets helps agents save time and increase productivity.
Suggested macro badges and confidence levels
Suggested macro badges and confidence levels enhance existing suggested macro functionality. A red badge over the macro menu brings the agent’s attention to macros that may help them solve the current ticket. Within the macro menu, confidence level indicators show how likely it is that the macro is relevant to the ticket the agent is working on.
Enhance writing
Enhance writing uses generative AI to improve the content of agent comments and help center content in different ways:
- Expand: Rephrases the content of an agent comment or selected help center content with additional language.
- Simplify: Rephrases the content of your comment to be more concise.
- Make more friendly: Changes the tone of an agent comment or selected help center content so that it sounds more conversational.
- Make more formal: Changes the tone of an agent comment or selected help center content so that it sounds more professional.
- Write a custom prompt: Changes your comment based on the custom instructions you enter that enhance or adjust the text.
- Simplify (help center only): Changes the selected content of a help center article to be clearer and more concise.
- Summary (help center only): Displays a concise description of a help center article’s content.
Tone of voice
Tone of voice uses generative AI to adapt agent responses to the overall writing style and tone of a conversation, and also allows auto assist to adapt its suggestions accordingly. The system rewrites or suggests text by analyzing certain conversation variables from past interactions between the end user and the agent within a given ticket. Some of the conversation variables that the system analyzes include:
- Overall tone and language
- Greetings, closures and signatures
- Writing style and language complexity (for example, length and format)
- Use of specific words/eponyms and special characters
- Ticket conversation sentiment, topic, criticality and urgency
Ticket summaries
Ticket summaries use generative AI to recap the public comments that have been added to a ticket so far, helping agents get up to speed and respond to customers more quickly. Internal notes can be included in the summary if needed.
Quick answers for Agent Workspace
Quick answers for Agent Workspace provide AI-generated answers to agents’ searches within Knowledge in the context panel. Agents can then copy the answer directly into the ticket or click the link to the source article that contains the generated answer.
Similar tickets
Similar tickets let agents see a list of resolved tickets that are similar to the one they’re currently working on. By seeing how similar issues were addressed, agents are able to solve tickets more efficiently.
Merging suggestions
Merging suggestions identifies tickets that might be potentially merged with the ticket the agent is currently working on. These suggestions appear in the context panel within a ticket.
Intelligent triage predictions
Intelligent triage predictions, including intent, language, and sentiment, give agents additional context about the ticket that helps them resolve customer requests quickly.
Entity detection
Entity detection allows you to define and detect unique information in tickets and messaging conversations. You can then use that information to populate associated custom ticket fields and power automated workflows and reports.
Creating procedures
Using a tool like Gemini or ChatGPT you can create your procedures from articles or pieces of text. Copy the procedure steps (step 1 to step 5) and ask the tool to create a Zendesk CoPilot procedure using those steps and the information you provided. If you have links in your text, ask the tool to include the links as well.
For a full tutorial on how to implement each of these features, please check out this YouTube video.
Comments
0 comments
Article is closed for comments.