Quick actions and shortcuts
You do not have to wait for the AI assistant to initiate things itself — you can control it directly via the chat field with simple commands and questions.Drafts and responses
| Type this | What happens |
| "Write a response" | The AI assistant generates a draft based on the ticket context |
| "Write a short response" | A shorter draft with only the essentials |
| "Write a formal response" | A draft with a more formal tone |
| "Write in English" | A draft in English regardless of the ticket language |
| "Try again" | A new draft — useful if the first one was not good |
| "Make it shorter / longer" | Rewrites the draft to the desired length |
| "Add an apology" | Adds empathy and an apology to the draft |
Research and context
| Type this | What happens |
| "What do we know about this customer?" | Retrieves ticket history and saved memory |
| "What does our help center say about X?" | Searches the knowledge base and summarises |
| "Has the customer had this problem before?" | Searches previous tickets |
| "What is our policy on X?" | Searches the knowledge base and instructions |
| "Summarise the ticket history" | Provides a short overview of what has happened |
Rules and memory
| Type this | What happens |
| "Show all rules" | Lists all active rules for the instance |
| "Save as a rule: [instruction]" | Saves a new rule (with confirmation) |
| "Remember that this customer [X]" | Saves a memory at the customer level |
| "Save for this ticket: [note]" | Saves a memory at the ticket level |
| "Save for this org: [note]" | Saves a memory at the org level |
| "What do we know about this organisation?" | Retrieves saved memory for the org |
Macros
| Type this | What happens |
| "Suggest a macro" | The AI assistant recommends the most relevant macro |
| "Is there a macro for X?" | Searches for a specific type of macro |
Tips for effective use
- Be specific — "Write a friendly response explaining that we are investigating the return request" is better than "Write a response"
- Give feedback in the chat — correct the AI assistant directly and save good corrections as rules
- Use quick actions for research — you do not need to leave the ticket view to find information
- Combine — "What is our returns policy, and then write a response to the customer based on it"