Privacy Policy

Effective 28 August 2026

This policy describes the preview version of MultiVoice Studio. The publisher identity and hosting region will be finalized before paid launch.

Information we process

We process account details, speaker-labelled scripts, voice reference recordings, generated narration, project settings, credit usage, billing identifiers supplied by Stripe, and security logs needed to operate the service.

How we use information

We use this information to authenticate you, save projects, create voice profiles and narration, account for usage, fulfil subscriptions, prevent abuse, troubleshoot failures, and meet legal obligations. We do not sell personal information.

Voice data

Voice references can identify or resemble a person. Upload a voice only when you are the speaker or have explicit permission to clone and use it. References and generated audio are private to your account. The launch architecture stores them on the service's persistent data volume; production deployment must add encrypted storage and backup controls before paid voice cloning is enabled.

AI-generated audio

Chatterbox-generated audio is synthetic and includes Resemble AI's PerTh watermark. The service also marks render responses as AI-generated. You are responsible for any additional disclosure required where you publish or play the audio.

Service providers

Stripe processes payment and billing data when enabled. The hosting provider processes application traffic and stored data. OpenAI processes text only when the optional OpenAI TTS engine is configured and selected. The current preview uses a hosted Chatterbox adapter that may send text and uploaded reference clips to the public Resemble AI Chatterbox Space; an optional Hugging Face token authenticates quota but does not make that Space private. A private model deployment will be configured before sensitive or paid voice data is accepted. Final subprocessor links will be added at commercial launch.

Security and retention

Passwords are hashed with scrypt. Session and OAuth tokens are stored as hashes. Access checks separate each account's projects, renders, and voices. No system is completely secure. Production launch requires TLS, managed secrets, encrypted backups, monitoring, and tested deletion and restore procedures.

How long we keep data

Generated narration audio is retained for 30 days and then automatically deleted; download anything you want to keep before then. Scripts, projects, and voice reference recordings are kept until you delete them in the studio or close your account. Deleting your account permanently removes your projects, voice profiles, renders, and personal data from the active service. Credit and billing records are retained as long as required for accounting and tax purposes.

Your choices

You can remove individual voice profiles from the studio, export account data from the Account panel, or permanently delete the account after password confirmation.

Contact

For privacy questions and data requests, contact hello@printya.sg.