Add Call Center Voice Agent with Azure Speech

[sazed-bot]

Source URLs

• Candidate source: https://github.com/Azure-Samples/call-center-voice-agent-accelerator
• GitHub source: https://github.com/Azure-Samples/call-center-voice-agent-accelerator

Assessor score

• Score: 5
• Version: assessor_llm_v2

Synopsis

Azure-Samples template for a real-time speech-to-speech call center voice agent built on Azure Voice Live API, with pluggable telephony via Twilio Voice, Azure Communication Services, or Infobip, plus a web browser test client. Ships as an azd-deployable accelerator (Bicep infra + Python server) targeting Azure Container Apps, with customizable prompts and optional ambient call-center audio scenes.

Reasoning

Repo is a real, MIT-licensed, publicly forkable accelerator with concrete azd deployment, multi-provider telephony, and explicit STT/LLM/TTS via Azure Voice Live. It clears all disqualifiers: not a single-vendor hello-world (composes telephony + Azure Voice Live + Container Apps), not opinion/marketing/walkthrough, has exportable artifact (repo + Bicep + server code). Outcome is moderately specific (call-center inbound agent) but not vertical-deep, so C3 gets a 3. Otherwise strong across schema, fork, and stack.

Claude raw output

{
  "id": "call-center-voice-agent-azure-speech",
  "slug": "call-center-voice-agent-azure-speech",
  "tags": [
    "call-center",
    "azure-speech",
    "telephony",
    "real-time",
    "multi-provider",
    "azure-communication-services",
    "twilio",
    "infobip"
  ],
  "title": "Call Center Voice Agent with Azure Speech",
  "config": {
    "features": [
      "real-time ASR",
      "real-time TTS",
      "ambient audio mixing",
      "multi-provider telephony",
      "Azure AI Foundry Agent integration"
    ],
    "ambient_scenes": [
      "none",
      "office",
      "call_center"
    ],
    "deployment_target": "Azure Container Apps",
    "telephony_providers": [
      "Azure Communication Services",
      "Twilio",
      "Infobip",
      "Web Browser"
    ]
  },
  "source": "github_discovery",
  "license": "MIT",
  "demo_url": null,
  "industry": "other",
  "pipeline": {
    "llm": "GPT-4o",
    "stt": "Azure Speech",
    "tts": "Azure Neural TTS",
    "telephony": "Twilio Voice"
  },
  "use_case": "support",
  "framework": "Custom",
  "languages": [
    "en-US"
  ],
  "created_at": "2026-05-25T19:30:00Z",
  "raw_prompt": null,
  "updated_at": "2026-05-25T19:30:00Z",
  "contributor": {
    "github": "Emma-ms"
  },
  "description": "A real-time speech-to-speech voice agent for call centers powered by Azure Voice Live API. Integrates with Azure Communication Services, Twilio, or Infobip for telephony, and includes a web browser client for testing. Supports customizable prompts, ambient audio scenes, and optional Azure AI Foundry Agent connections.",
  "prompt_file": null,
  "github_source_url": "https://github.com/Azure-Samples/call-center-voice-agent-accelerator"
}

Dedup decision

F6 dedup has not landed yet; no dedup decision available.

Reviewer notes

approve-to-pr running: