What the software does
The grind, done in minutes.
It listens to the whole recording, transcribes it, and marks every long pause, stray “um” and promising moment. It never tires at minute fifty. It never cuts anything on its own.
Roll 01The editor’s amplification systemPodcasts & video
One-button AI editors cut the editor out of the edit, and their mistakes ship with your name on them. We went the other way. The machine does the grind: every pause, every filler, every moment worth clipping, found in minutes. You make every call. Then you finish it right here, with captions, vertical framing and a real timeline. One room. Your judgment, at machine speed.
Where we stand on AI
What the software does
It listens to the whole recording, transcribes it, and marks every long pause, stray “um” and promising moment. It never tires at minute fifty. It never cuts anything on its own.
What you do
You hear each proposal in context and cut it, keep it or move it, in minutes instead of hours. You know which silence is dead air and which one is the point. Nothing is applied until you build it.
Judgment is an editor's job. We built the machine to make you faster at it, and that's why your edits hold up where fully automatic ones fall apart.
Let's talk about the fear
If you edit for a living
They flatten the pause before the punchline. They clip the setup and miss the payoff. They can't tell a filler from a beat. Judgment is the one part of editing that doesn't automate, so we automated everything except judgment. You keep the craft. We take the hours.
If your team is unsure about AI
Nothing here is cut, captioned or framed on its own. Every proposal waits for a person to say yes, every decision can be reversed, and your recording stays private to your account. The machine works for your editor, and your editor answers to you.
Editors who use AI won't be replaced by AI. They'll be the ones setting the pace. Be that editor.
Scroll to run the edit. Tap an orange mark to keep that pause instead.
Illustrated example.
The Amplifier
One room for the whole edit. Clean it up, rearrange it, frame it for vertical, caption it, add titles and your own music, then build it once. Undo anything. Every keystroke an editor expects.
Illustrated example.
Two ways to win with one recording
Mode one
One finished episode, audio or video. Pauses longer than a breath are shortened, not deleted, so people still sound like people. Hesitation sounds that sit on their own are removed. Volume is leveled to the podcast standard.
Mode two
The moments worth sharing, as separate files of up to 90 seconds. They are picked by what is said, not sliced every minute and a half: a complete story, a strong opinion, a useful tip. Each one has to make sense on its own.
Why us
How it works
Drop, paste or pick a file up to 5 GB. MP3, WAV, M4A, FLAC, OGG, AIFF, MP4, MOV, WEBM or MKV.
The recording is transcribed and scanned. Edits and clips are proposed. Nothing is cut yet.
Cut or keep each proposal, then arrange, reframe and caption in the editing room. Every decision is saved and can be reversed.
Play it in your studio. If something's off, reopen the editor with your decisions intact. One re-edit is included.
Download the files. Your original is never altered.
The standard
Other tools hand you whatever the algorithm produced. Here, nothing is cut without your say, and every file is held to six checks before it reaches you.
Pricing
Founding price
$50 per recording
Up to 4 hours of audio or video, in either mode. Every cut is yours to make, and one re-edit is included.
Start an episodeEvery order includes
Your recording, protected
Files are never at a public address. Downloads use links that expire in 15 minutes and only work while you are signed in.
Only you review your recording. Transcription and drafting run on Cloudflare's AI models. The AI proposes; it never applies a cut on its own.
Originals are deleted 30 days after delivery and finished files after 90. You can delete everything yourself at any time.
Progress, files and our replies all live in your studio. Nothing gets lost in an inbox.
Your studio shows the exact step your recording is on, from upload to delivery.
Titles and notes are written from your full transcript and checked so no name or number appears that wasn't in the recording.
Questions
Audio or video up to 5 GB and 4 hours: MP3, WAV, M4A, FLAC, OGG, AIFF, MP4, MOV, WEBM, MKV. Drag it in, paste it, or pick it from your device or cloud drive app.
No. Pauses under about a second and a half are left alone, and longer ones are shortened rather than removed. Only hesitation sounds that stand apart are cut. Phrases such as “you know” are flagged for you and kept unless you decide otherwise.
The full transcript is read and moments are nominated for being complete and interesting on their own. Each is trimmed to clean sentence boundaries, capped at 90 seconds, and handed to you to play, trim and pick. The number depends on how much of the recording is worth sharing.
Clean edit: the edited episode (MP3, or MP4 for video), a transcript, title options, show notes and the list of edits. Highlight clips: each clip as its own file with a title.
One re-edit is included. Reopen the editor, change what you need, and it's rebuilt from your original. Your other decisions stay as you left them.
Message us from the contact page. Replies appear right there in your studio.
Upload a recording today. Let the machine do the grind, make every call yourself, and ship a finished episode or a stack of vertical, captioned clips the same day.