Data Processing
Speechall routes audio and applicable options to the transcription provider for the model you select. The model catalog identifies providers and shows which models support file transcription and streaming.
Speechall infrastructure
Cloudflare Pages serves the website. The API deployment is configured for Fly.io in Ashburn, Virginia, and the Speechall database is configured for Supabase in Frankfurt, Germany. Locations for audio processing, provider support, and provider backups depend on the selected provider and its configuration.
File transcription can use temporary processing storage. Speechall schedules cleanup for those audio files and keeps operational records used for usage, billing, and support. See the Privacy Policy for the types of information those records can contain.
Provider retention, training, and deletion
Handling can differ by provider, endpoint, and account configuration. Speechall requests use the provider terms that apply to Speechall’s account. Before sending sensitive audio, request written confirmation of retention periods, training use, processing locations, transfer safeguards, and deletion terms for the models you plan to use.
Procurement and data requests
To request a data processing agreement, the subprocessor schedule, or provider-specific information, email [email protected]. Include the model IDs, whether you need file transcription or streaming, and your regional and retention requirements. For an existing job, include its identifier when available. Keep audio, transcripts, and API keys out of your message.