Native Voice Agent Support using Pipecat

[Aakash326]

Feature Area

Core functionality

Is your feature request related to a an existing bug? Please link it here.

no

Describe the solution you'd like

I would like to propose adding native voice agent support to CrewAI through integration with Pipecat.

CrewAI provides excellent multi-agent orchestration, task management, memory, and tool execution. However, building real-time voice agents currently requires significant custom integration work.

Pipecat is an open-source framework focused on real-time voice and multimodal agents, providing:

• Speech-to-Speech pipelines
• Streaming STT/TTS
• Interruptions and barge-in handling
• Real-time transports (Daily, Twilio, WebRTC)
• Voice agent workflows

Motivation

Many production AI systems are moving toward voice-first experiences:

• Customer support agents
• Sales agents
• Scheduling assistants
• Internal enterprise assistants

Currently, developers must manually connect CrewAI with voice frameworks.

A native integration could allow CrewAI agents to operate as real-time voice agents while keeping CrewAI's existing strengths around:

• Agent collaboration
• Task orchestration
• Memory
• Tool usage

Potential Design

Example API:

voice_crew = VoiceCrew(
    crew=my_crew,
    transport="daily",
    stt="deepgram",
    tts="cartesia"
)

Potential MVP:

• Pipecat transport adapter
• Streaming user input to CrewAI agents
• Streaming CrewAI responses back to users
• Support for interruptions
• Shared memory across voice interactions

Questions

1. Is native voice support aligned with CrewAI's roadmap?
2. Would the maintainers be open to a Pipecat-based integration?
3. If so, what would be the preferred architecture before implementation?

Happy to discuss and contribute an MVP if this direction is interesting.

Describe alternatives you've considered

No response

Additional context

No response

Willingness to Contribute

Yes, I'd be happy to submit a pull request