Skip to content
Microsoft Teams Bridge for ElevenLabs Agents

Microsoft Teams Bridge for ElevenLabs Agents

Welcome! @komaa/elevenlabs-msteams-bridge (npm: @komaa/elevenlabs-msteams-bridge) puts an ElevenLabs Agent on a real Microsoft Teams call.

It is a small Node.js service (and importable TypeScript library) that sits between two WebSockets:

flowchart LR
    Teams["Microsoft Teams<br/>call"]
    StandIn["StandIn media bridge<br/>(hosted service)"]
    Bridge["this bridge<br/>(yours)"]
    EL["ElevenLabs Agent<br/>STT + LLM + TTS + VAD"]
    Teams <--> StandIn
    StandIn -- "HMAC WS" --> Bridge
    Bridge -- "WS" --> EL

The hosted StandIn media bridge (standin.komaa.com) joins the Teams call and dials into your bridge, one authenticated WebSocket per call. The bridge opens one ElevenLabs Agent conversation per call and relays between them. Both sides speak base64 PCM 16 kHz mono, so audio is copied verbatim in both directions: no resampling, no re-encoding, nothing added to the latency budget beyond a relay hop.

What it does

Section titled “What it does”
  • Realtime voice - the caller talks to your ElevenLabs agent and hears it reply. Turn-taking, VAD and interruption are the agent’s own; the bridge maps the caller’s barge-in onto the Teams side and drops stale “ghost” audio so nothing plays after an interruption.
  • Per-call personalization - caller name, tenant and direction as dynamic_variables; optional localized greeting or spoken AI disclosure; per-caller memory via user_id (guests never share an identity).
  • Vision on demand - the agent’s look client tool sees the caller’s camera or screen-share: your own vision model (path 2) or ElevenLabs multimodal upload (path 1, recording-gated).
  • Agent client tools - end_call, express (avatar emotion), show_image (image on the bot’s tile, SSRF-guarded), look.
  • Two call governors - StandIn-side tier cutoffs get a spoken goodbye; a bridge-side MAX_CALL_MINUTES hard cap protects your ElevenLabs budget.
  • Hardened transport - replay-proof HMAC handshake, connection and payload caps, duplicate-call rejection, graceful SIGTERM drain, and an ElevenLabs host allowlist so your API key can never be sent elsewhere.

Use the sidebar to navigate. Start with Getting Started, or jump to the Configuration Reference or Wire Protocol. There is also a runnable example project in the repository.