Sumnote: Sub-processors

Last updated: July 13, 2026

This page lists the third-party service providers (“sub-processors”) that process personal information on Sumnote’s behalf to run the Service. It supplements our Privacy Policy.

We do not sell your personal information. We share it only with sub-processors who process it on our behalf, under terms that require them to protect it and use it only for the services they provide to us. We maintain this list current and give customers advance notice (at least 30 days where practicable) before adding a new one.

Always-on sub-processors

These process data for every customer.

ProviderPurposeWhat they processLocation
Amazon Web ServicesCloud infrastructure and storageAll customer data (account, meeting audio, transcripts, files)United States
AnthropicAI extraction of action items, decisions, summariesMeeting transcripts (text only)United States
OpenAIBackup speech-to-text transcription; backup extractionMeeting audio (transcription); transcripts (extraction)United States
AssemblyAIPrimary speech-to-text transcriptionMeeting audioUnited States
RunPodGPU compute for training Sumnote's own modelsThe redacted training set only (identifiers stripped; no audio; no voiceprints)United States
Google (Gmail SMTP)Outbound emailRecipient email address and message contentUnited States
Google (Firebase Cloud Messaging)Push notificationsDevice push tokens and notification previewsUnited States

Conditional sub-processors

These process data only in the circumstances stated.

ProviderActivates whenWhat they process
Google (Workspace APIs)You sign in with Google or connect Google Calendar / MeetAuthentication tokens; calendar events you sync; meeting metadata
ZoomYou connect your Zoom accountAuthentication tokens and meeting metadata. The call itself is handled by Zoom as your meeting platform, under Zoom's own terms
StripeWe launch paid plans and you subscribeBilling details and transaction history (card data handled by Stripe directly)

Note. We use AssemblyAI for speech-to-text only; voiceprints are computed within our own systems, not by AssemblyAI. We do not use any analytics provider that tracks you inside the application.