Advanced Agent Settings
6. Advanced Agent Settings
Advanced Settings allow you to configure optional features that affect how the agent behaves during conversations and the overall user experience.
You can enable or disable the following features according to your requirements.
| Feature | What It Provides |
|---|---|
| Memory | Remembers previous conversations |
| Follow-Up Questions | Suggests questions to continue the conversation |
| Response Editing | Saves predefined responses |
| Quotes and Citations | Displays source references |
| Speech-to-Text | Supports voice input |
6.1 Memory
Memory determines whether the agent remembers previous conversations.
When this feature is enabled, the agent can remember previous conversations and maintain conversational context. This allows users to continue a conversation without having to explain previously discussed topics again in every message.
You can enable Memory for long-running conversations where maintaining context is important.
6.2 Follow-Up Questions
Follow-Up Questions allows the agent to generate suggested questions that users can ask to continue the conversation.
When this feature is enabled, the agent can suggest follow-up questions after its response to help continue the conversation or explore the topic in greater depth.
This makes it easier for users to see what they can ask next and continue interacting with the agent.
6.3 Response Editing
Response Editing allows you to manually save high-quality responses that the agent can prioritize when responding to similar user questions.
This feature can be used when you want more controlled, predefined responses to be provided for specific questions.
6.4 Quotes and Citations
Quotes and Citations allows the sources and referenced sections used in the agent's responses to be displayed to users.
When this feature is enabled, users can see which sources the information provided by the agent is based on. It is particularly useful for agents that work with documents and knowledge sources, allowing users to trace the sources behind responses.
6.5 Speech-to-Text
Speech-to-Text allows users to communicate with the agent using voice input.
When this feature is enabled, voice input provided by the user is converted into text and sent to the agent. This allows users to create messages using their voice instead of typing them.
7. Describing an Agent: Notes and Suggested Prompts
These fields allow you to document the purpose of your agent and make it easier for users to start interacting with it.
7.1 Notes
The Notes field allows you to enter a short description of the agent or explain its intended purpose.
This field is particularly useful when you have multiple agents, as it helps you quickly understand what each agent was created for.
For example:
An assistant that interprets data tables, provides insights, and accelerates decision-support processes.
7.2 Initial Suggested Prompts
Initial Suggested Prompts allow you to define ready-to-use questions or request examples that users will see when starting a new conversation with the agent.
By adding prompts that match the agent's intended purpose, you can show users how they can interact with the agent.
For example, for a data analysis agent:
What can I do with this agent?
Can you analyze the data I added?
Can you summarize the significant changes in the data?
These prompts help users understand the agent's use cases more quickly and make it easier to start a conversation.