# EchoEnvo Sub-processors

Version 2026-11-05 · Draft, not yet in effect

These providers process personal data for EchoEnvo. We tell customers who
signed our [Data Processing Addendum](/legal/dpa.md) about a new or replaced provider
at least 30 days in advance, by email and on this page, and they can object.

| Provider | What it does for us | Data | Location | Transfer safeguard |
|---|---|---|---|---|
| OpenAI | speech recognition, translation and editing of transcripts, text checks | audio of speech, transcripts, translations | United States | {{OPENAI_SAFEGUARD}} |
| Fish Audio | speech synthesis, voice cloning | translated text, voice samples, voice models | United States | {{FISH_SAFEGUARD}} |
| Stripe | payments, tax calculation, fraud prevention | billing details, card data, payment history | United States, EU, UK | EU–US DPF and UK extension; SCCs |
| Cloudflare | network delivery, security, country of visitor | IP address, request data | global | EU–US DPF and UK extension; SCCs |
| Google Workspace | email from EchoEnvo and support mailbox | email address, message content | United States, EU | EU–US DPF and UK extension; SCCs |
| {{HOSTING_PROVIDER}} | servers, database, file storage | all data in the service | {{HOSTING_REGION}} | {{HOSTING_SAFEGUARD}} |
| {{GPU_PROVIDER}} | separating voice from background, aligning words in time | audio | {{GPU_REGION}} | {{GPU_SAFEGUARD}} |

**Fish Audio and model training.** At the date of this list, Fish Audio's
public terms allow it to use uploaded content and usage data to develop and
train its models. EchoEnvo itself does not train models on customer data. We
are seeking a contract with Fish Audio that excludes training on EchoEnvo data
and will update this list when it is in place.
