@millions-hair-24956 how to transition naturally through appended prompts based on the context of the transcription?

It seems likely that transitioning naturally through appended prompts based on transcription context can be managed by using the

before_llm_cb

callback in LiveKit's VoicePipelineAgent. This callback allows you to customize the reply by injecting context or additional information before the LLM synthesizes a response. You can append prompts or modify the chat context dynamically based on the transcription received. Would you like more detail on how to implement this? Sources: - https://livekit-users.slack.com/archives/C07FY8WHGPM/p1733426789.731689 - https://livekit-users.slack.com/archives/C07FY8WHGPM/p1731640235.491609 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.