Added a new section on configuring the AI tester. - #1172
Conversation
|
🌿 Preview your docs: https://vapi-preview-01a04430-3935-7448-a475-453430636e7c.docs.buildwithfern.com |
|
🌿 Preview your docs: https://vapi-preview-01a04480-6d06-75b9-94bd-97266ef499ae.docs.buildwithfern.com |
scott-lowe-vapi
left a comment
scott-lowe-vapi
left a comment
There was a problem hiding this comment.
Nice addition. The dashboard flow and remaining API examples match current behavior. I left three focused comments.
| -H "Authorization: Bearer $VAPI_API_KEY" | ||
| ``` | ||
|
|
||
| Then update the personality with `PATCH`. The following request shows representative settings for the model, Soniox transcriber, Vapi Voice, and conversation behavior: |
There was a problem hiding this comment.
Built-in personalities can be fetched here, but the API rejects PATCH updates to them. Since the quickstart sends users from a default personality to this page, could we add a branch: POST an org-owned copy when orgId is null, otherwise PATCH?
| | **Transcriber** | Provider, model, and language | Language detection and hints, endpoint delay, vocabulary, contextual hints, and fallback transcribers | | ||
| | **Voice** | Provider and voice ID | Speed, version, language, pronunciation dictionaries, chunking, formatting, and caching | | ||
|
|
||
| Transcriber and voice settings apply to voice simulations. Chat simulations use the model only. |
There was a problem hiding this comment.
Chat simulations also honor firstMessage and who starts first. Could we say: ‘Chat simulations ignore transcriber and voice settings; behavior, model, and turn-taking still apply.’
| }, | ||
| "stopSpeakingPlan": { | ||
| "numWords": 2, | ||
| "voiceSeconds": 0.2, |
There was a problem hiding this comment.
voiceSeconds is only used when numWords is 0, so it has no effect in this example. Could we remove it or set numWords to 0?
|
🌿 Preview your docs: https://vapi-preview-01a044d8-248b-76a8-a908-0840a518b207.docs.buildwithfern.com |
- Use 'assistant or squad under test' instead of 'target-assistant' - Map Behavior/Model/Transcriber/Voice settings to the Personality tab - Bridge the Dashboard 'intent' field to the API instructions field - Use 'success criteria' consistently in the cURL section Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
|
🌿 Preview your docs: https://vapi-preview-01a044e9-252d-757a-9090-d2f5d43beb86.docs.buildwithfern.com |
|
🌿 Preview your docs: https://vapi-preview-01a04512-cbc9-749e-adf6-644c7cff92bf.docs.buildwithfern.com |
2 participants
Description
Testing Steps
fern docs devor navigate to preview deployment