Last updated May 24, 2026
To deliver VoiceTextTranslate we share limited data with the following third-party providers. Each operates under a data processing agreement that restricts them to processing personal data on our documented instructions.
This page is the authoritative list. When we add, remove, or change a sub-processor, we update this page and bump the "Last updated" date. If you use VoiceTextTranslate in a context where sub-processor changes matter to you, check back periodically.
| Provider | Role | Data processed | Location | Links |
|---|---|---|---|---|
| Stripe | Payment processing | Card details (collected directly by Stripe, never touch our servers), billing email, transaction history. | United States | Site Privacy DPA |
| Deepgram | Speech-to-text | Audio streamed from users for real-time transcription, and audio files uploaded for async transcription. Transcripts returned to our server. | United States | Site Privacy DPA |
| Gladia | Speech-to-text and translation | Audio streamed from users for real-time speech-to-text and translation. Audio files for async transcription. Translated text output. | European Union (France) | Site Privacy DPA |
| DeepL | Text translation | Source text sent for translation and the translated output returned to our server. Used for async project translation. | European Union (Germany) | Site Privacy DPA |
| Google Cloud | Text-to-speech synthesis | Text sent to Google Cloud Text-to-Speech API for voice synthesis. Audio returned to our server; not retained by Google long-term. | United States (with regional data residency options) | Site Privacy DPA |
| DigitalOcean | Application hosting, database, and object storage | All structured data processed by the VoiceTextTranslate application: user account data, authentication tokens, usage records, purchase history, project metadata, and request/response payloads. Also all user-uploaded files (audio, video), derived artifacts (transcripts, translations, normalized audio, subtitle exports), and downloadable batch export archives — stored in DigitalOcean Spaces. | European Union (FRA1, Frankfurt) — droplet, managed Postgres, and Spaces bucket all colocated | Site Privacy DPA |
| Amazon Web Services | Compute for video subtitle burn-in (AWS Fargate) | Video files and subtitle tracks are read from DigitalOcean Spaces into ephemeral Fargate task memory for subtitle burn-in, then the rendered result is written back to DigitalOcean Spaces. No persistent storage on the AWS side; Fargate tasks exit after each job. | European Union (eu-central-1, Frankfurt) | Site Privacy DPA |
| MailerSend | Transactional email delivery | Recipient email addresses, subject lines, and email body content for invitations, account confirmations, password resets, and export-ready notifications. | European Union (Lithuania) | Site Privacy DPA |
| Sentry | Error reporting and application monitoring | Application exception reports, which may incidentally include user email addresses, IP addresses, and request payloads when an error surfaces on a user-initiated action. | United States | Site Privacy DPA |
| Grafana Cloud | Observability (traces, logs, metrics) | OpenTelemetry traces, logs, and metrics from the Rails application, inference gateway, and supporting infrastructure containers. PII is redacted at our edge collector before transmission (emails, Bearer tokens, sensitive query/form params) — see Telemetry::Redactor for the patterns. May still incidentally include request paths, opaque user identifiers, IP addresses, and error context captured by automatic instrumentation. | European Union (Ireland, eu-west-2) | Site Privacy DPA |
For more information about how we handle your personal data, see our Privacy Policy.