Record with consent
Tell participants what is being recorded, why, who can access it and how long you plan to retain it. Upload only material your agreement and local rules allow you to process.
For product managers and UX researchers
Turn a saved user interview into a transcript you can inspect
Move from an interview recording to searchable, editable text without losing the original conversation. Find a phrase, correct a name, label the speakers and copy a verified quote into your research workspace.
A practical interview-to-evidence workflow
Transcribe the saved file
Upload audio or record in the app. Automatic language detection is available across 20 supported languages, including right-to-left reading for Hebrew, Arabic, Farsi and Yiddish.
Review speakers and terms
Add product names and domain terms under “Words and names to recognize,” then review the result. Speaker separation is optional and not perfect: rename a detected speaker, reassign an individual turn, and check consequential wording against the recording.
Move verified text onward
Search the library, copy selected text, download a TXT file, or create an expiring transcript link. Keep your analysis and participant context in the research system your team already uses.
Built around the work after the call
For freelance product managers
A searchable transcript reduces the time spent scrubbing through audio when you need to recover a customer phrase, confirm an objection or prepare evidence for a client readout. A revocable share link can give a client access without sending the recording around.
For UX researchers
Keep the participant's words available while you tag observations, compare sessions and develop themes elsewhere. Review sensitive passages before sharing, and use participant codes rather than names when your research plan calls for pseudonymisation.
For multilingual interviews
The product transcribes the language that was spoken; it does not turn one language into another. That boundary preserves a source-language record while leaving translation decisions with your team or a qualified language professional.
What you still need to do
Transcripts does not generate research findings, themes, summaries or product recommendations. It does not know whether a participant is representative, whether a quote is safe to publish, or how the interview fits the rest of your evidence. Treat the text as a navigable draft of the recording: verify consequential quotes against the audio, apply your consent and retention policy, then perform synthesis with the context only your research team has.
Questions before you upload
Can it identify the interviewer and participant?
Automatic speaker separation can organise detected voices into turns. It can make mistakes, especially with overlap or poor audio, so labels are editable and attribution should be checked before a quote leaves the research team.
Can a client review the result?
You can create an expiring transcript link and revoke it later. Confirm that sharing is permitted, remove unnecessary identifiers, and choose the audience before sending any participant material.
Does it analyse the interview?
No. The output is a transcript in the spoken language. Coding, synthesis, themes, decisions and recommendations remain separate research work.
Start with one saved interview
Eligible new customers receive a 7-day, card-required trial with a one-time 120-minute transcription allowance — about 60 audio minutes with automatic speaker separation. After that, the subscription is $8.99/month. Returning customers are charged immediately.
Public checkout is not open yet. Request early access to test the interview workflow.
Third-party AI providers transcribe the audio, Cloudflare R2 stores media, and recordings stay in your private account library until you delete them. Review the privacy page before processing participant data.