The ElevenLabs integration brings AI voice generation directly into a workflow's Voice Over stage, as an alternative to routing to a human voice artist.

What it's used for

At a Voice Over stage, a completed script can be sent to ElevenLabs to generate a spoken audio file automatically — useful for fast-turnaround spots, draft versions for internal review before a final human recording, or languages where a suitable human voice artist isn't immediately available.

Connecting an API key

An admin adds an ElevenLabs API key under the tenant's integration settings. Once connected, the integration is available to any workflow template's Voice Over-type stages — there's no per-template setup beyond the one-time tenant-level key.

Configuring a stage to use it

On a Voice Over (Task-type) stage, the assignment type can be set to route to the ElevenLabs integration instead of a human role. The stage configuration lets you select a voice profile and language, matched against the ticket's own language field the same way a human voice artist would be skill-matched (see Skills & Languages).

ElevenLabs-generated audio is still tracked through the same media library and revision history as a human-recorded file — if a script changes and the audio needs regenerating, the new version is stored as a new iteration, not an overwrite.

Next: Storage Providers covers where the resulting files actually get stored.