Pressentifylegal
Sub-processors
Last updated 31 October 2026
When a company runs interviews on Pressentify, the company decides what happens to a candidate's data and we process it on its behalf. These are the services we use to do that. Companies are told 30 days before a service is added or replaced. This list is a draft while the company-hiring terms are being reviewed.
The list
- MongoDB Atlas — Database. Data: Accounts, invitations, answers, marks, reviews, audit rows, wrapped recording keys. Region: Cloud region chosen at setup (to be stated per customer). Their terms
- Cloudflare (R2 storage, DNS) — File storage and DNS. Data: Encrypted interview recordings (ciphertext only), other uploaded files, database backups (encrypted). Region: Automatic; EU/US placement set per bucket. Their terms
- Vercel — Application hosting (until the move to our own server). Data: Requests and responses of the web application; no recording passes through it. Region: To be stated per customer. Their terms
- OpenRouter and the model providers it routes to — AI marking and interviewer lines. Data: Answer text and the organisation's rubric for each marking call; no video, no audio, no name. Region: Provider dependent. Their terms
- NVIDIA (API catalogue) — AI marking and interviewer lines (alternative provider). Data: Answer text and the organisation's rubric for each marking call; no video, no audio, no name. Region: United States. Their terms
- LiveKit — Media relay for live interviews (the media service). Data: Live audio and video in transit between interviewer and candidate; not stored by us at LiveKit. Region: LiveKit Cloud region chosen at setup. Their terms
- Email provider (Gmail SMTP) — Invitations, reminders and confirmations. Data: Candidate email address and the text of the message. Region: Provider dependent. Their terms
Questions and changes
Write to pressentify@gmail.com to ask about a service or to be told when the list changes.