ellai.com/build/conversation-flow/finetune-examples) [(4) ]() .

### Answer Fine-tune examples in Conversation Flow help guide the agent's behavior by providing sample transcripts as context [(1) ]() . You configure them by providing a transcript where you select the role (

user

,

agent

, or

function

) for each part . When selecting

function

as the role, you can fill out both the invocation and result of the function . Where fine-tune examples are supported: - **Conversation Node**: Supports fine-tune examples for both response generation and transition logic [(1) ]() [(2) ]() - **Function Node**: Supports fine-tune examples for transition [(1) ]() - **Extract Dynamic Variable Node**: Supports fine-tune examples for transition [(3) ]() How to configure them: For conversation response fine-tuning, you only need to provide the transcript as context [(1) ]() . It's not necessary to provide the entire call transcript—you can simply provide the relevant part . At least one

agent

response is required since this is fine-tuning the agent's response . For transition fine-tuning, you need to provide both a transcript as context and the transition result . If you cannot distinguish between different nodes available as transition targets, you can rename your nodes to make them easier to distinguish . The sources don't explicitly detail the internal injection mechanism, but the fine-tune examples work by showing the model what kind of responses or transitions you expect in specific conversational contexts [(1) ](

​ To learn more, see the sources I considered: [[1] Finetune Examples](https://docs.retellai.com/build/conversation-flow/finetune-examples) [[2] Conversation Node](https://docs.retellai.com/build/conversation-flow/conversation-node) [[3] Extract Dynamic Variable Node](https://docs.retellai.com/build/conversation-flow/extract-dv-node) [[4] Debug guide](https://docs.retellai.com/build/conversation-flow/debug-guide) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.