Build a voice agent with LangChain

from typing import AsyncIterator
import asyncio
from assemblyai_stt import AssemblyAISTT
from events import VoiceAgentEvent

async def stt_stream(
    audio_stream: AsyncIterator[bytes],
) -> AsyncIterator[VoiceAgentEvent]:
    """
    Transform stream: Audio (Bytes) → Voice Events (VoiceAgentEvent)

    Uses a producer-consumer pattern where:
    - Producer: Reads audio chunks and sends them to AssemblyAI
    - Consumer: Receives transcription events from AssemblyAI
    """
    stt = AssemblyAISTT(sample_rate=16000)

    async def send_audio():
        """Background task that pumps audio chunks to AssemblyAI."""
        try:
            async for audio_chunk in audio_stream:
                await stt.send_audio(audio_chunk)
        finally:
            # Signal completion when audio stream ends
            await stt.close()

    # Launch audio sending in background
    send_task = asyncio.create_task(send_audio())

    try:
        # Receive and yield transcription events as they arrive
        async for event in stt.receive_events():
            yield event
    finally:
        # Cleanup
        with contextlib.suppress(asyncio.CancelledError):
            send_task.cancel()
            await send_task
        await stt.close()

from langchain_core.utils.uuid import uuid7
from langchain.agents import create_agent
from langchain.messages import HumanMessage
from langgraph.checkpoint.memory import InMemorySaver

# Define agent tools
def add_to_order(item: str, quantity: int) -> str:
    """Add an item to the customer's sandwich order."""
    return f"Added {quantity} x {item} to the order."

def confirm_order(order_summary: str) -> str:
    """Confirm the final order with the customer."""
    return f"Order confirmed: {order_summary}. Sending to kitchen."

# Create agent with tools and memory
agent = create_agent(
    model="google_genai:gemini-3.1-pro-preview",  # Select your model
    tools=[add_to_order, confirm_order],
    system_prompt="""You are a helpful sandwich shop assistant.
    Your goal is to take the user's order. Be concise and friendly.
    Do NOT use emojis, special characters, or markdown.
    Your responses will be read by a text-to-speech engine.""",
    checkpointer=InMemorySaver(),
)

async def agent_stream(
    event_stream: AsyncIterator[VoiceAgentEvent],
) -> AsyncIterator[VoiceAgentEvent]:
    """
    Transform stream: Voice Events → Voice Events (with Agent Responses)

    Passes through all upstream events and adds agent_chunk events
    when processing STT transcripts.
    """
    # Generate unique thread ID for conversation memory
    thread_id = str(uuid7())

    async for event in event_stream:
        # Pass through all upstream events
        yield event

        # Process final transcripts through the agent
        if event.type == "stt_output":
            # Stream agent response with conversation context
            stream = agent.astream(
                {"messages": [HumanMessage(content=event.transcript)]},
                {"configurable": {"thread_id": thread_id}},
                stream_mode="messages",
            )

            # Yield agent response chunks as they arrive
            async for message, _ in stream:
                if message.text:
                    yield AgentChunkEvent.create(message.text)

from cartesia_tts import CartesiaTTS
from utils import merge_async_iters

async def tts_stream(
    event_stream: AsyncIterator[VoiceAgentEvent],
) -> AsyncIterator[VoiceAgentEvent]:
    """
    Transform stream: Voice Events → Voice Events (with Audio)

    Merges two concurrent streams:
    1. process_upstream(): passes through events and sends text to Cartesia
    2. tts.receive_events(): yields audio chunks from Cartesia
    """
    tts = CartesiaTTS()

    async def process_upstream() -> AsyncIterator[VoiceAgentEvent]:
        """Process upstream events and send agent text to Cartesia."""
        async for event in event_stream:
            # Pass through all events
            yield event
            # Send agent text to Cartesia for synthesis
            if event.type == "agent_chunk":
                await tts.send_text(event.text)

    try:
        # Merge upstream events with TTS audio events
        # Both streams run concurrently
        async for event in merge_async_iters(
            process_upstream(),
            tts.receive_events()
        ):
            yield event
    finally:
        await tts.close()

export LANGSMITH_TRACING="true"
export LANGSMITH_API_KEY="..."

import getpass
import os

os.environ["LANGSMITH_TRACING"] = "true"
os.environ["LANGSMITH_API_KEY"] = getpass.getpass()

from langchain_core.runnables import RunnableGenerator

pipeline = (
    RunnableGenerator(stt_stream)      # Audio → STT events
    | RunnableGenerator(agent_stream)  # STT events → Agent events
    | RunnableGenerator(tts_stream)    # Agent events → TTS audio
)

# Use in WebSocket endpoint
@app.websocket("/ws")
async def websocket_endpoint(websocket: WebSocket):
    await websocket.accept()

    async def websocket_audio_stream():
        """Yield audio bytes from WebSocket."""
        while True:
            data = await websocket.receive_bytes()
            yield data

    # Transform audio through pipeline
    output_stream = pipeline.atransform(websocket_audio_stream())

    # Send TTS audio back to client
    async for event in output_stream:
        if event.type == "tts_chunk":
            await websocket.send_bytes(event.audio)