Quick Replies
Overview
Quick Replies (QR) allow AI Colleagues to send clickable buttons as part of bot responses, enabling users to take actions with a single tap instead of typing. This improves user experience, reduces friction, and streamlines task completion in conversational flows. When enabled, the LLM powering the AI Colleague intelligently decides when to include Quick Reply buttons based on the conversation context—no manual configuration is needed per response.
Configuration
Administrators can enable or disable Quick Replies from the Settings → Orchestrator Settings → Enable quick reply buttons.
| Setting | Description |
|---|---|
| Enable Quick Replies | Master toggle to allow AI Colleagues to send Quick Reply buttons in bot responses |
How it works
The decision to display Quick Reply buttons is made automatically by the LLM based on the following criteria:
| Criteria | Description |
|---|---|
| Response context | Is this an actionable response that requires the user to make a choice? |
| User query type | Is the user asking for options, a menu, or help? |
| Workflow stage | Is this an approval step, confirmation step, or decision point in a process? |
Administrators do not need to configure individual Quick Reply buttons per skill or AOP. The LLM evaluates the conversation in real time and attaches appropriate buttons when applicable.
Use Cases
Quick Replies are particularly effective in the following scenarios:
- Approval notifications — Present Approve / Reject buttons when a task or request requires user approval, enabling one-tap action directly in the chat.
- Help and menu queries — When a user asks "What can you do?" or "Menu," the AI Colleague can surface relevant capability options as Quick Reply buttons for easy navigation.
- Confirmation flows — Display Yes / No or Confirm / Cancel buttons during workflows that require explicit user confirmation before proceeding.
- Navigation — Provide options such as Go Back, Main Menu, or Related Topics to help users move through multi-step conversations without typing.
Quick Reference
| Scenario | Default |
|---|---|
| Approval / rejection steps | ✓ Yes |
| Help, menu, or capability queries | ✓ Yes |
| Confirmation prompts (Yes/No) | ✓ Yes |
| Navigation options (Go Back, Main Menu) | ✓ Yes |
| Simple knowledge base responses | ✗ Typically no |
| Mid-process / intermediate messages | ✗ Typically no |
| Clarification questions | ✓ Yes. For option based inputs |
Note: Since Quick Reply display is LLM-driven, the above table represents typical behaviour. The LLM may adapt based on conversational context.
Updated about 16 hours ago