Pipecat logo

Agent Orchestration

Pipecat

Official Python transport adapter (pipecat-agentduet) linking AgentDuet phone and WhatsApp calls into modular Pipecat pipelines with STT, LLM, TTS, and zero webhook hosting.

Voice frameworkPipecatPipelinesVADInterruption handlingZero-webhook
Documentation
Where it fits

Pipecat in your voice stack.

  1. Pipecat - declarative voice pipelines

    TelephonyPhone networkInbound call
    AgentDuetAgentDuet TransportPersistent outbound connection
    Speech-to-speechLive modelListens, call functions, replies
    ToolsActionsCRM, calendar, APIs
Overview

What it contributes to the stack.

Pipecat is an open-source framework for building modular, real-time voice and multimodal conversational pipelines.

pipecat-agentduet is the official Python transport adapter that bridges AgentDuet telephony directly into Pipecat. It provides `AgentDuetTransport`, streaming live two-way 16-bit PCM phone and WhatsApp audio through your STT, LLM, TTS, and VAD services.

Because the AgentDuet SDK connects outbound to the platform, you do not need to host public webhook endpoints, run FastAPI servers, configure ngrok tunnels, or transcode audio formats.

For setup, credentials, and API details, use the Documentation link above.

Capabilities

The technical characteristics that matter when deciding whether this provider belongs in your application.

Zero webhook hosting

Connects outbound from your bot process without public IPs, tunnels, or webhook infrastructure.

Declarative pipelines

Wire Deepgram, Google Gemini, OpenAI, Cartesia, ElevenLabs, and Silero VAD in concise pipelines.

Native PCM streaming

Streams mono 16-bit PCM directly at the call rate (8 kHz, 16 kHz, or 24 kHz) with no manual transcoding.

Automated call lifecycle

Answers on pipeline start, cleans up on completion, and cancels automatically on caller hangup.

Common use cases

Practical scenarios where the provider's role is clear and the surrounding systems remain under application control.

Inbound AI receptionist

Greet incoming callers, run real-time conversational flows, and handle interruptions seamlessly.

Modular voice AI stacks

Swap STT, LLM, or TTS providers across different numbers without rewriting telephony logic.

Custom VAD & turn processing

Combine Silero VAD and custom turn processors with live AgentDuet phone audio.

Related integrations

Compare nearby choices.

These providers occupy a similar role in the stack. Compare their strengths, operating model, and surrounding services.

Google ADK

Google's Agent Development Kit for live agent sessions with tools, specialists, and call-scoped state.

Agent frameworkGoogleMulti-agentToolsLive sessions
Explore integration

LangGraph

LangGraph for durable agent workflows: graphs, checkpoints, and policy-gated tools around live phone sessions.

Agent frameworkStatefulWorkflowsDurableGraph orchestration
Explore integration

Gemini Live

Google's real-time speech model for live phone agents. Stream call audio in and play spoken replies back in the same session.

Speech-to-speechMultimodalInterruptionTool callingNative audio
Explore integration
Browse every integration

Build a voice agent.

AgentDuet handles the phone and messaging boundary. Your model runs the conversation.

Explore the SDK