Back to Projects
TypeScript

voice-agent-starter-kit

Self-hosted, multilingual phone voice agent starter kit (LiveKit + Sarvam), ~700ms-1.2s latency. ⭐ 28 stars on GitHub.

View Source Code

System Overview

What the project does

An open-source starter kit for building a self-hosted, multilingual phone voice agent, generalized from a production Voice AI stack.

Key features

  • Real-time voice pipeline built on LiveKit for low-latency audio streaming
  • Sarvam for multilingual speech (Indian languages) STT/TTS
  • ~700ms-1.2s round-trip latency, tuned for natural phone conversations
  • Self-hostable end-to-end, no vendor lock-in
  • Sanitized reference architecture ready to fork and adapt

    • Tech stack

    LiveKit, Sarvam AI, TypeScript/Node.js, WebRTC.

    Use case

    Teams that want to stand up a phone-based multilingual voice agent (support, sales, IVR replacement) without building the real-time audio pipeline from scratch.

    Architecture Details

    This system integrates multiple components for a seamless automation flow. Structural interpretation based on project focus:

    Backend Infrastructure

    Core execution layer for robust data processing and API handling.

    AI / Logic Core

    Intelligent decisioning via models or logical workflow rules.

    Tech Stack

    TypeScriptIntegrationAutomationAPIs

    Key Capabilities

    • ▹ Custom workflow execution
    • ▹ Data transformation and routing
    • ▹ Extensible architecture