# # Copyright (c) 2024–2025, Daily # # SPDX-License-Identifier: BSD 2-Clause License # import os from bedrock_agentcore import BedrockAgentCoreApp from dotenv import load_dotenv from loguru import logger from pipecat.audio.vad.silero import SileroVADAnalyzer from pipecat.frames.frames import LLMRunFrame from pipecat.pipeline.pipeline import Pipeline from pipecat.pipeline.runner import PipelineRunner from pipecat.pipeline.task import PipelineParams, PipelineTask from pipecat.processors.aggregators.llm_context import LLMContext from pipecat.processors.aggregators.llm_response_universal import LLMContextAggregatorPair from pipecat.processors.frameworks.rtvi import RTVIConfig, RTVIObserver, RTVIProcessor from pipecat.runner.types import RunnerArguments, SmallWebRTCRunnerArguments from pipecat.runner.utils import create_transport from pipecat.services.aws import AWSBedrockLLMService from pipecat.services.cartesia.tts import CartesiaTTSService from pipecat.services.deepgram.stt import DeepgramSTTService from pipecat.transports.base_transport import BaseTransport, TransportParams from pipecat.transports.smallwebrtc.connection import IceServer, SmallWebRTCConnection from pipecat.transports.smallwebrtc.request_handler import ( IceCandidate, SmallWebRTCPatchRequest, SmallWebRTCRequest, SmallWebRTCRequestHandler, ) from pipecat.transports.smallwebrtc.transport import SmallWebRTCTransport app = BedrockAgentCoreApp() request_handler: SmallWebRTCRequestHandler = None load_dotenv(override=True) # We store functions so objects (e.g. SileroVADAnalyzer) don't get # instantiated. The function will be called when the desired transport gets # selected. transport_params = { "webrtc": lambda: TransportParams( audio_in_enabled=True, audio_out_enabled=True, vad_analyzer=SileroVADAnalyzer(), ), } async def run_bot(transport: BaseTransport, runner_args: RunnerArguments): logger.info(f"Starting bot") yield {"status": "initializing bot"} stt = DeepgramSTTService(api_key=os.getenv("DEEPGRAM_API_KEY")) tts = CartesiaTTSService( api_key=os.getenv("CARTESIA_API_KEY"), voice_id="71a7ad14-091c-4e8e-a314-022ece01c121", # British Reading Lady ) # Automatically uses AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, and AWS_REGION env vars. llm = AWSBedrockLLMService( model="us.amazon.nova-2-lite-v1:0", params=AWSBedrockLLMService.InputParams(temperature=0.8), ) messages = [ { "role": "system", "content": "You are a helpful LLM in a WebRTC call. Your goal is to demonstrate your capabilities in a succinct way. Your output will be spoken aloud, so avoid special characters that can't easily be spoken, such as emojis or bullet points. Respond to what the user said in a creative and helpful way.", }, {"role": "user", "content": "Say hello and briefly introduce yourself."}, ] context = LLMContext(messages) context_aggregator = LLMContextAggregatorPair(context) rtvi = RTVIProcessor(config=RTVIConfig(config=[])) pipeline = Pipeline( [ transport.input(), rtvi, stt, context_aggregator.user(), llm, tts, transport.output(), context_aggregator.assistant(), ] ) task = PipelineTask( pipeline, params=PipelineParams( enable_metrics=True, enable_usage_metrics=True, ), idle_timeout_secs=runner_args.pipeline_idle_timeout_secs, observers=[RTVIObserver(rtvi)], ) @rtvi.event_handler("on_client_ready") async def on_client_ready(rtvi): logger.info(f"Client ready") await rtvi.set_bot_ready() # Kick off the conversation. await task.queue_frames([LLMRunFrame()]) @transport.event_handler("on_client_connected") async def on_client_connected(transport, client): logger.info(f"Client connected") @transport.event_handler("on_client_disconnected") async def on_client_disconnected(transport, client): logger.info(f"Client disconnected") await task.cancel() runner = PipelineRunner(handle_sigint=runner_args.handle_sigint) task_id = app.add_async_task("voice_agent") await runner.run(task) app.complete_async_task(task_id) yield {"status": "completed"} async def initialize_connection_and_run_bot(request: SmallWebRTCRequest): """Handle initial WebRTC connection setup and run the bot.""" raw_urls = os.getenv("ICE_SERVER_URLS") urls = [u.strip() for u in raw_urls.split(",") if u.strip()] ice_servers = [ IceServer( urls=urls, username=os.getenv("ICE_SERVER_USERNAME"), credential=os.getenv("ICE_SERVER_CREDENTIAL"), ) ] transport = None runner_args = None async def webrtc_connection_callback(connection: SmallWebRTCConnection): nonlocal transport, runner_args runner_args = SmallWebRTCRunnerArguments( webrtc_connection=connection, body=request.request_data ) transport = await create_transport(runner_args, transport_params) yield {"status": "initializing connection"} global request_handler request_handler = SmallWebRTCRequestHandler(ice_servers=ice_servers) answer = await request_handler.handle_web_request( request=request, webrtc_connection_callback=webrtc_connection_callback ) yield {"status": "ANSWER:START"} yield {"answer": answer} yield {"status": "ANSWER:END"} async for result in run_bot(transport, runner_args): yield result async def add_ice_candidates(patch_request: SmallWebRTCPatchRequest): """Handle ICE candidate additions for existing connections.""" await request_handler.handle_patch_request(patch_request) yield {"status": "success"} @app.entrypoint async def agentcore_bot(payload, context): """Bot entry point for running on Amazon Bedrock AgentCore Runtime.""" request_type = payload.get("type", "unknown") logger.info(f"Received request of type: {request_type}") data = payload.get("data") if not data: logger.error("No data found in payload") yield {"status": "error", "message": "No data found in payload"} return match request_type: case "offer": # Initial connection setup try: request = SmallWebRTCRequest.from_dict(data) except Exception as e: logger.error(f"Failed to deserialize SmallWebRTCRequest: {e}") yield {"status": "error", "message": f"Invalid request payload: {str(e)}"} return async for result in initialize_connection_and_run_bot(request): yield result case "ice-candidates": # ICE candidate additions try: if "candidates" in data: data["candidates"] = [IceCandidate(**c) for c in data["candidates"]] patch_request = SmallWebRTCPatchRequest(**data) except Exception as e: logger.error(f"Failed to deserialize SmallWebRTCPatchRequest: {e}") yield {"status": "error", "message": f"Invalid request payload: {str(e)}"} return async for result in add_ice_candidates(patch_request): yield result case _: logger.error(f"Unknown request type: {request_type}") yield {"status": "error", "message": f"Unknown request type: {request_type}"} return # Used for local development async def bot(runner_args: RunnerArguments): """Bot entry point for running locally and on Pipecat Cloud.""" transport = await create_transport(runner_args, transport_params) async for result in run_bot(transport, runner_args): pass # Consume the stream if __name__ == "__main__": # NOTE: ideally we shouldn't have to branch for local dev vs AgentCore, but # local AgentCore container-based dev doesn't seem to be working, or at # least not for this project. if os.getenv("PIPECAT_LOCAL_DEV") == "1": # Running locally from pipecat.runner.run import main main() else: # Running on AgentCore Runtime app.run()