ElevenLabs integration

Separate voice quality from agent behavior.

When your agent uses ElevenLabs voices, Word Is Bond scores the conversation outcome while still exposing latency and transcript accuracy.

Setup

Expose the conversational agent through a target you control.
Use expected lines to track scripted voice-agent behavior.
Run barge-in tests to see if the agent handles interruption.
Compare runs after voice/model/prompt changes.

Target config

{ "targetAgent": { "transport": "direct", "peerId": "elevenlabs-agent" } }

CI example

curl -X POST https://api.wordis-bond.com/api/tests/run \
  -H "Authorization: Bearer $WORDIS_BOND_KEY" \
  -H "Content-Type: application/json" \
  -d '{"targetAgent":{"transport":"direct","peerId":"elevenlabs-agent"},"goal":"Answer pricing questions and book a follow-up","bargeIn":true}'

Test your ElevenLabs agent this afternoon

Free tier, no credit card. Point Word Is Bond at an agent you run and read the four-layer scorecard.