The Conversation Settings section allows you to customize how your AI Agent interacts with users throughout a conversation. You can configure voice-specific behaviour, enable contextual memory, define inactivity handling, and manage message processing settings to create a seamless conversational experience.

From this section, you can:

• Configure Voice Settings for voice-enabled AI Agents, including speech profiles, interruption handling, speech processing, and hold behaviour.
• Enable Contextual Memory to allow the AI Agent to leverage information from previous customer interactions.
• Configure Timeout Settings to determine how the AI Agent responds when users do not reply within a specified period.
• Manage Message Controls, including response streaming, filler messages, multiple user messages, and conversation reset behaviour.

These settings help tailor the AI Agent's behaviour to your business requirements and deliver more personalised and efficient customer interactions.

Voice Settings​

For voice-enabled AI Agents, you can configure the following settings:

Speech Profile

Select the speech profile used by the Voice AI Agent. A speech profile is required for all voice-enabled AI Agents.

Barge-in Functionality

Control whether users can interrupt the AI Agent while it is speaking.

When enabled, you can configure:

• Sensitivity: Defines the speech completion threshold required before an interruption is recognised.
• Time Prompt: Specifies the minimum duration the AI Agent must speak before interruption is allowed.
• Barge-in Response Time: Determines how long the system waits after detecting user speech before processing the interruption.

The system uses the highest applicable value when deciding when interruptions can occur.

Speech End Threshold

Defines the duration of silence after a user finishes speaking before the AI Agent begins processing the response.

Global Hold Settings

Enable hold handling for voice conversations. When enabled, the configured hold workflow is triggered whenever a hold intent is detected.

Available settings include:

• Hold Intent: Select the intent that triggers the hold workflow.
• User Reply Wait Time: Specify how long the AI Agent waits for a user response while on hold.
• Hold Response: Define the message sent when a hold intent is detected.
• Enable Conditions: Configure conditions that determine when a hold response is used.
• Barge-in Functionality: Allow users to interrupt the hold message.
• Voice Reply: Configure the voice message played when a hold intent is detected.
• + Add Voice Reply: Add multiple hold responses for different scenarios.

Default Values

Unless modified, the following default values apply:

​

​

Contextual Memory

Contextual Memory enables the AI Agent to retain and use relevant customer information across conversations, helping deliver more personalised and context-aware interactions.

Timeout Settings

Timeout Settings determine how the AI Agent handles situations where a user becomes inactive.

• Enable Timeout Settings: Activates timeout handling.
• User Reply Wait Time: Specifies how long the AI Agent waits before a timeout occurs.
• Timeout Response: Defines one or more responses sent when a timeout is triggered.

For each timeout response, you can configure:

• Enable Conditions to determine when the response should be used.
• Barge-in Functionality to control whether users can interrupt the timeout message.
• Voice Reply to configure the message played after a timeout.

Additional settings include:

• + Add Voice Reply: Add multiple timeout responses for different scenarios.
• Repeat Count: Specify how many times the AI Agent should repeat the timeout response.
• Timeout Task: Configure the fallback action that executes after all timeout attempts have been exhausted.

Message Controls

The Message Controls section allows you to configure additional messaging behaviours for your AI Agent, helping deliver a more responsive and seamless conversational experience.

Stream Response

Enable Stream Response to send AI-generated responses to users in real time as they are generated, rather than waiting for the complete response to be ready.

​

Filler Message

Displays a temporary message while the AI Agent processes a user request.

Filler Prompt

Specify the message shown to users while the AI Agent generates or processes a response.

Multiple User Messages

Control whether users can send additional messages while the AI Agent is processing a previous request.

• Enabled: Users can send multiple messages during processing.
• Disabled: Users must wait for the AI Agent to respond before sending another message.

Reset Context Keywords

Configure keywords that automatically reset the current conversation context. When a user enters one of the configured keywords, the AI Agent starts a new conversation and treats the message as the first user input.

Common examples include:

• Hi
• Reset

Additional Settings

Configure the following settings to manage model behavior and compliance:

• Select Guardrail: Apply guardrails to user inputs and AI Agent responses to help enforce compliance and safe interactions.
• PII Masking Template: Select the template used to identify and mask personally identifiable information (PII) within conversations.
• Select LLM Provider: Choose the large language model (LLM) provider used by the AI Agent.
• Model: Select the model that generates AI Agent responses.

​

Advanced LLM Settings

Enable Additional Settings to configure the following model parameters:

• Max Tokens: Specifies the maximum number of tokens allowed across the request and generated response.
• Temperature: Controls response variability. Lower values produce more predictable responses, while higher values generate more diverse outputs.
• Frequency Penalty: Reduces repetitive content by penalizing frequently used tokens.
• Presence Penalty: Encourages the model to introduce new concepts by penalizing tokens that have already appeared in the response.

​