Nothing retained
Audio is streamed for transcription and never stored on our servers — a one-line data-handling story that ethics committees and participants can actually understand.
Interviews, ethnographic sessions, user research — transcribed as they happen, so the gap between fieldwork and analysis is minutes, not a week of typing.
An hour of audio is famously four to six hours of typing, and outsourcing it means another consent hurdle, a budget line, and a wait. Meanwhile the detail you remember about session three fades before the transcript arrives.
Transcribing live closes that gap: you leave the session with the verbatim text already done, and your field notes can point at timestamps instead of paraphrases.
Used by academic researchers, UX researchers and evaluation teams.
Audio is streamed for transcription and never stored on our servers — a one-line data-handling story that ethics committees and participants can actually understand.
Line-level timestamps mean your codes can reference the exact moment, and checking a quote against the audio is a click, not a scrub.
You get what was said, attributed to who said it — no AI paraphrase deciding what mattered before you’ve done the analysis.
The data-handling is simple to describe: audio is relayed for transcription during the session and never stored afterwards; only the transcript remains, and it stays yours. Check the specifics against your protocol — the privacy policy states exactly what is processed.
Yes — the transcript is what was said, speaker by speaker, not a summary. Cleaning up false starts is your editorial call, made in your copy of the text.
Mixed-language speech is one of the things the underlying model is built for, and it’s a common reality of field interviews. Errors still cluster at language switches, so review those passages.