What we keep, and why.
Last updated: 26 April 2026. Plain-English summary, not a legal document — see the Terms for the lawyer-grade version.
What we collect
When you run a session: your interview transcript, any email you give us, a hashed device fingerprint (so we can rate-limit free sessions without asking you for an account), and basic technical metadata about your browser and connection.
We do not collect: your name, address, phone number, or anything we don't need to deliver the report.
What we store and for how long
Audio is processed in-flight for transcription and immediately discarded. Your transcripts stay encrypted at rest in our database for 30 days on the free tier and as long as you keep your account on paid plans.
Your reports are kept for the same window as your transcripts. You can delete everything from your account page (or by emailing us) at any time.
Who we share it with
For the voice loop to work we send transcripts to Anthropic (model inference) and audio chunks to Deepgram (speech-to-text) and Google Cloud Text-to-Speech. None of these providers retain your data for training, per their enterprise terms.
For email delivery we use Resend. For abuse protection we use Cloudflare Turnstile. For rate-limit storage we use Upstash Redis. Your evaluation reports live on Neon Postgres.
We do not sell your data. We do not run advertising trackers.
Your rights
You can request a copy of your data, ask us to delete it, or object to any processing — at any time, no reason needed.
Email us at privacy@interviewmentor.ai.