Pipeline with lookup example

voice-agent-example.py

async def start_query_knowledge_base(function_name, llm, context):
    """Push a frame to the TTS service; this is handy when the LLM response might take a while."""
    await llm.push_frame(TTSSpeakFrame("Please hold on while I look that order up for you."))

async def query_knowledge_base(function_name, tool_call_id, args, llm, context, result_callback):
  query_result = knowledge_base.get(args["query"])
  await result_callback({
    "info": json.dumps({
      "lookup_success": True,
      "order_status": query_result["order_status"],
      "delivery_date": query_result["delivery_date"],
    })
  })

llm.register_function("query_knowledge_base",
  query_knowledge_base
  start_callback=start_query_knowledge_base
)

pipeline = VoiceAgentPipeline(
  [
      transport.input(),               # Audio in from user
      stt,                             # Speech to text
      context_aggregator.user(),       # Manage 'user' context
      llm,                             # LLM inference
      tts,                             # Text to speech
      transport.output(),              # Audio out to Network
      context_aggregator.assistant(),  # Manage 'assistant' context
  ]
)