Q&A
Q. When did you know the first version wasn't working?
The day support tickets outran signups. That was the tell — people wanted it, they just couldn't get through the door.
Q. What changed?
We deleted the onboarding. All of it.
Interview transcription for journalists
Record the conversation, not the notes. Interviews come back with speakers told apart and a verbatim transcript you can check any quote against — and every interview after comes back the same way, with a Q&A and pull quotes when you want them.
The day support tickets outran signups. That was the tell — people wanted it, they just couldn't get through the door.
We deleted the onboarding. All of it.
For journalists
The transcript is the floor — these are the documents on top, every one yours to adjust.
Verbatim transcript
Word for word, speakers told apart — the record you quote from.
Interview → Q&A
The conversation, cleaned into questions and answers.
Pull quotes
The lines worth running, verbatim, with who said them.
The one you define
There’s no menu — describe your document once and it’s a type like any other.
The proof — a real recording, the real documents
Speaker labels tell interviewer from interviewee; the verbatim transcript sits beside every document, so a quote you run is a quote they said.
The day support tickets outran signups. That was the tell — people wanted it, they just couldn't get through the door.
We deleted the onboarding. All of it.
The mechanism
01
Record and upload
Phone, field recorder or a call — upload the file or paste a link when you're back at your desk.
02
It becomes the verbatim record
Word for word, speakers told apart, timestamps kept.
03
Your documents write themselves
Q&A, pull quotes, or your own format — checked against the transcript before you file.
A correction you confirm applies from the next interview on — and you can undo any of them.
Conduct, not badges
Deleted fast, provably, never fed to a model to learn from — and never sold. The verbatim transcript is the record; check any quote against it before you run it.
Hours are the only meter.
Documents, corrections and exports are never counted — you choose hours, everything else is included. There’s a free tier, no card.
The short answers
Yes — speaker identification labels interviewer and interviewee automatically, so attribution is never a guess. Rename the labels once and they stick.
Yes. Click any line of the transcript to hear that exact moment in the recording — verify the quote before you publish it.
Yes — full-text search across your transcripts finds that one line about policy changes without scrubbing back through hours of audio.
By default they're deleted after 7 days — keep a project longer when a story needs it — every deletion writes you a receipt, and nothing you record is used to train AI models.