Voice operations control

Keep every campaign clean, synced, and easy to manage.

Review client lists, correct voice assignments, upload new assets, and publish changes to the allocator from one workspace.

1 unified workspace
GCS voice folder validation
Audit every change recorded
Quick filtering by root, variant, and voice name
Cleaner add, delete, and upload workflows
Better at-a-glance campaign context before you edit
Secure sign in

Welcome back

Enter your account details to open the voice dashboard.

ID
Key

Protected Authenticated access only
Live Connected to campaign data and storage
VP

Voice Pipeline

Campaign Workspace

Navigate

Workspace actions

Current workspace

Dashboard

Review voice coverage, filter variants, and inspect the active campaign.

Select a campaign table to begin.

Signed in as
⚠
Vox voice missing
Active campaign

No table selected

Choose a campaign table to load voices, variants, and audit history.

0 tables Waiting for data
Client roots 0

Distinct root groups currently available in this table.

Voice rows 0

Total loaded rows available for filtering and editing.

Focus Dashboard

Section currently optimized for the task in front of you.

Overview

Voice library health

Filter by client root and variant, then inspect the current voice mix before editing.

No dataset loaded

Filters

Focus the dataset before making changes.

Storage reference

Cross-check campaign voices available in GCS.

0 voices

Global voice lookup

Search for a voice name across all client lists to see where it is whitelisted or blacklisted. Type a single capital letter (e.g. N) to list every voice of that flow.

Client ID Voice Type

Total voices 0

Rows in the current filtered scope.

Whitelist 0

Approved voices in the selected view.

Blacklist 0

Blocked voices in the selected view.

Pending deletions

Voices scheduled for removal — they stay live until the countdown ends. Cancel to keep them.

Client ID Voice Requested by Source Executes in Cancel

Voices in selected scope

Use filters above to narrow the result set.

Client ID Voice Type Actions
Build coverage

Add or create voices

Use the mode switch below to extend an existing list or spin up a new client list.

Add to existing variant

Select a root and variant, then paste one voice name per line.

Create a new list

Start a new client root and seed it with valid voices in one action.

Operator notes

Before you add

  • Voice names are validated against available GCS folders.
  • Formatting is enforced, so malformed voice names are rejected early.
  • Use the dashboard first if you need to confirm whether a variant already exists.
Cleanup

Remove voices safely

Delete specific voices or remove an entire list variant when it is no longer needed.

Delete selected voices

Filter down to a single client variant, then select the voices to remove.

Select a client root and variant first.

Delete an entire list

Remove every voice inside a selected variant. Use this only when the list is obsolete.

This will delete all voices in the selected list. This action is immediate.
Destructive actions

Work carefully

  • Delete individual voices when you only need to remove a few entries.
  • Delete the full list only after you confirm the variant is no longer used.
  • The audit log can help you verify recent changes after cleanup.
Storage

Upload a voice folder

Send a voice set to Google Cloud Storage with a cleaner file selection flow.

Prepare upload

Select the target campaign, set the voice name, then upload files or a folder.

CSV with a Tag column. Click Check CSV Tags to compare against the simulator without uploading any audio, or upload it together with a voice folder to get both reports at once.
Uploading a Vox11 voice manually? Enter its History ID, click Verify ID to confirm it matches, and the link will be saved and reported to the portal after upload. Voices added via the Import button below are linked automatically.
Select voice folder(s)

Click Add Folder to select a voice folder. Repeat to add more folders.

Or use Select Files to pick individual audio files.

Each folder name becomes the voice name. Path: Production/{campaign}/{folder}/All/

Selected voices (0 folders, 0 files)

Audio processing

Applied automatically to .raw files before upload.

Upload guidance

Processing and naming

  • Trim removes leading and trailing silence below -40 dB, keeping 20 ms padding.
  • Normalize scales peak amplitude to -1 dB below full scale for consistent loudness.
  • Only .raw files (16-bit PCM, 8 kHz) are processed. Other file types are uploaded as-is.
  • Dialogue Flow — if set, uploaded files are checked against the audio tags API. Upload is blocked if any tags are missing.
  • Max Turn — controls how many dialogue turns to check for required tags.
  • Folder structure — each voice folder must contain an All/ subfolder with the audio files (e.g. EmilyA/All/*.raw).
  • After upload, return to the dashboard to verify the voice appears in storage lookups.

Import from Vox11

Click Import to pull a reviewed voice from the Vox11 portal into the upload form above — it appears as a selected folder. Then set the campaign, dialogue flow, and processing options and click Upload to GCS as usual. The portal is notified automatically after a successful upload.

The voice folder name defaults to the ZIP name (e.g. LenaDefaultA, last letter = flow) — you can change it when importing.

History ID Voice Folder Vox11 Voice Campaign Flow Variant Tags Reviewed Status Action
Click "Load Generations" to fetch reviewed voices from Vox11.
History

Audit trail

Review recent activity for the currently selected table and confirm operational changes.

Timestamp Action User Details
Inline edit

Edit voice entry

Confirm delete

Delete this voice?

Are you sure you want to delete this voice?