Privacy by design.

You keep the transcripts. We delete the audio. SpeechWeave deletes source media after processing and never uses customer audio or transcripts to train any AI models.

Retention modes: standard keeps transcript text until TTL; metadata-only wipes transcript text after successful webhook delivery; zero-persistence is early access only.

API traffic uses TLS in transit. Staged source media is encrypted at rest in object storage until deletion. Operated by New Chapter America LLC. Enterprise DPA requests: [email protected].