polyshSubprocessors
Subprocessors
Polysh uses a small number of specialist providers to host your workspace and run each processing step. Each one receives only what its step needs.
Last updated:
A subprocessor is a third party that processes personal data on our behalf when we provide Polysh to your organization. This list forms part of our Data Processing Addendum.
Current subprocessors
| Provider | Purpose | Data it receives | Location |
|---|---|---|---|
| Amazon Web Services, Inc. | Application hosting, database, job processing, logs, transactional email (Amazon SES), and some text-to-speech (Amazon Polly) | All workspace data and account identifiers; guest email addresses and sign-in codes for share links; narration text | European Union (Frankfurt, Germany) |
| Cloudflare, Inc. | Media storage (R2), upload service, website and web app hosting, DNS, and network security | Recordings, generated audio, exports, guide screenshots; IP addresses and request data | Global network |
| Clerk, Inc. | Sign-in, sessions, SSO, invitations, workspace membership, and the early-access waitlist | Names, email addresses, workspace roles, session and device data | United States |
| Deepgram, Inc. | Speech-to-text, language detection, narration timing, and text-to-speech | Recording audio, narration text and audio, team word-list terms | United States |
| OpenAI, L.L.C. | Script rewriting, translation, and planning and checking step-by-step guides | Transcripts, scripts, translations, team word lists, guide screenshots | United States |
| Sarvam AI Private Limited | Text-to-speech, transliteration, speech-to-text, and language detection for Indian languages | Narration text; audio from recordings in Indian languages | India |
| Google LLC | Web fonts in the web app, and our team email for support conversations | IP address and browser details when the app loads; the content of emails you send us | United States |
Our core application, database, and job processing run on AWS in the European Union. Some processing steps are performed by providers in the United States and India, as shown above.
How we choose and use them
- Only what each step needs. Speech-to-text gets audio, the language model gets text and guide screenshots, and voice providers get the approved script.
- No training on your content. We use API settings that keep your content out of provider model training. Requests to OpenAI are made with storage turned off, and we opt out of Deepgram's model improvement programme.
- Contracts and safeguards. Each provider is bound by a written agreement with confidentiality, security, and data protection terms, and, where data leaves the European Economic Area or the United Kingdom, a recognised transfer mechanism such as the EU Standard Contractual Clauses.
- Encryption. Data is encrypted in transit between Polysh and every provider, and at rest in our database and storage.
Our free browser tools on polysh.io run on your device and don't send your files to any of these providers.
Changes and objections
We'll update this page at least 30 days before a new subprocessor starts processing customer data, except in an emergency needed to keep the Service running securely, when we'll update it as soon as we can. To be told by email when this list changes, write to [email protected] with the subject "Subprocessor updates".
If you have a reasonable data protection objection to a new subprocessor, tell us within those 30 days. We'll work with you on a solution, and if we can't find one, you may end the affected part of the Service and receive a refund of any prepaid fees for it.