Overview
Note: This feature is exclusively available for Agentic workflows and supports the Voice channel only for now.
Conversation Fillers allow AI Agents to send short messages to contacts during voice calls when a response is taking longer than expected, especially during Workflows, API or MCP calls. Instead of silence that highlights this latency, contacts receive timely messages that reassure them the system is active and still processing their request, reducing the perceived latency while creating a more natural conversation flow.
The feature is available in two modes:
- Static Fillers: Admins define a pool of custom, predefined messages and control timing parameters. The AI Agent randomly selects from the available messages when a filler is needed.
- Dynamic Contextual Fillers: The system automatically generates context-aware filler messages in the contact's language, based on the active conversation and Skill. No message authoring is required.
Conversation Fillers are available on all accounts and apply to the Voice channel only. We are planning to make them available in more channels soon. The feature must be manually enabled for each AI Agent individually.
Note: We recommend testing Conversation Fillers in your environment before enabling them in production to ensure the messages and timing can be fine-tuned to your use case.
Static Fillers
Enable Static Conversation Fillers
- Open the AI Agent you want to configure.
- Open the Orchestrator Configurator, using the Settings button
- Navigate to the Conversation Fillers section.
- Enable the Conversation Fillers toggle and select Static fillers
Configure Timing Settings
Once enabled, configure the following timing fields:
- Channel: select Voice channels
- In the Initial delay field, enter the number of seconds before the first filler is sent. The allowed range is 1 to 5 seconds.
- In the Filler interval field, enter the number of seconds between each follow-up filler. The allowed range is 2 to 10 seconds.
Note: All timing and retry values must be whole numbers. Decimal values are not supported. For example, enter 2 instead of 2,5.
- In the Number of retries field, enter the maximum number of fillers to send while the AI Agent is still processing. The allowed range is 1 to 5.
Note: Once the maximum number of retries is reached, no additional fillers will be sent to the contact for that interaction.
Configure Filler Messages
You can configure up to 10 Initial Messages and up to 10 Retry Messages. When a filler is triggered, the AI Agent randomly selects one message from the available pool.
Initial filler
- Under Initial Messages, click Add a message.
- Enter a short, neutral message to be sent when the AI Agent first takes longer than expected to respond.
- Repeat for up to 10 messages.
Note: Initial Messages are sent only once per interaction, immediately after the initial delay threshold is reached. Keep these short and neutral. Example: "Ok, hold on a moment."
Follow-up Filler
- Under follow-up filler, click Add message to one or more messages. The AI Agent randomly selects one message from the available pool.
- Enter a short message indicating the system is still working on the request.
- Repeat for up to 10 messages.
Note: Retry Messages are sent during subsequent filler intervals while the AI Agent continues processing. These should indicate the AI agent is still active. Example: "Still working on it."
Note: You need to insert text for both the initial and follow-up filler to be able to save the configuration.
Save Your Configuration
Once all timing settings and messages are configured, click Save.
Dynamic Contextual Fillers
Overview
Dynamic Contextual Fillers automatically generate filler messages without requiring admins to write any copy. The system detects when an AI Agent's response is taking longer than at least 1 second during a voice call and sends a contextually relevant message to the contact.
Messages are:
- Generated in the same language as the ongoing conversation
- Tailored to the active Skill (Integration or Workflow) and conversation context
- Designed to make sense within the flow of the interaction — for example, "Let me check that for you..." when retrieving information, or "Still working on it..." when processing a request
Default behavior:
| Parameter | Value |
| Initial delay | Between 1 to 5 seconds |
| Filler interval | Between 2 to 10 after the end of the initial filler |
| Maximum retries | 5 |
Notes:
If the AI Agent's response becomes available while a filler is playing, the current filler completes before the AI Agent response is delivered. Fillers and responses will not overlap.
If no response is generated within the retry window, the system's standard timeout process will apply.
Enable Dynamic Contextual Fillers
- Open the AI Agent you want to configure.
- Navigate to the Conversation Fillers section.
- Enable the Dynamic Contextual Fillers toggle.
- Set the initial filler interval
- Set the follow-up filler interval and number of retries
- Click Save.
Frequently Asked Questions
Can I use both Static Fillers and Dynamic Contextual Fillers on the same AI Agent? No. You must choose one mode per AI Agent. Enable either Static Fillers or Dynamic Contextual Fillers, depending on the level of control and customization you need.
Do timing fields accept decimal values? No. All timing and retry values must be whole numbers. For example, enter 2 instead of 2.5.
Does this feature work on digital or text channels? No. Conversation Fillers are currently available for the Voice channel only. Support for additional channels is planned for a future release.
Do I need to enable Conversation Fillers for all my AI Agents at once? No. The feature must be enabled and configured individually for each AI Agent. This allows you to tailor the settings to each agent's specific use case.
What happens if the AI Agent responds while a filler message is being played? The current filler message completes first. The AI Agent's response is then delivered immediately after, with no overlap.
What happens if the AI Agent doesn't respond within the retry window? Once the maximum number of retries is reached, no additional fillers are sent. The system's standard timeout process applies.
Can Dynamic Contextual Fillers generate messages in any language? Dynamic Contextual Fillers generate messages in the same language as the ongoing conversation, using the conversation's language context automatically.
Is this feature available to all accounts? Yes. Conversation Fillers are available on all accounts but must be manually enabled per AI Agent.