Privacy Policy
Last updated: 2026-05-06
What we collect
When you use CaptionTwin we receive: (a) audio you record from your microphone, streamed to our transcription provider; (b) the resulting text transcripts and AI-generated notes; (c) usage counters such as total recording time; (d) a random anonymous identifier stored in your browser's localStorage so we can show you your own past sessions. If you upgrade to a paid plan, we additionally receive your email address and a Stripe customer ID.
How we use it
Audio is streamed in real time to Deepgram for transcription and is never stored on our servers — only the resulting text transcript is. When you finish a session, the transcript and AI-generated notes are saved to our Postgres database (Supabase, US East). We use Google Vertex AI (Gemini) to translate transcripts and generate study notes. We may use PostHog to analyze aggregate usage patterns.
Who we share with
We share data only with subprocessors strictly necessary to operate the service: Deepgram (live transcription), Google Cloud / Vertex AI (translation and notes), Supabase (database hosting), Stripe (payments), Vercel (hosting), and PostHog (analytics). We do not sell your data.
Your rights
You can delete any session at any time from the session detail page; deletion is immediate and removes the transcript, notes, and any stored audio. To delete your account entirely, email hello@captiontwin.com.
Cookies
We do not set advertising cookies. We use the browser's localStorage for an anonymous user identifier and Stripe sets its own cookies during checkout.
Contact
Questions? Email hello@captiontwin.com.