Any recording, turned into a clean transcript you can actually read

Hand your assistant an audio or video file and get back a tidy, timestamped transcript with each speaker labeled. No fiddling with tools, no scrubbing through the recording to find the one line you needed.

What it does

• Takes almost any recording you have: a phone voice memo, a call, an interview, a webinar export, a video file.
• Produces a clean transcript with timestamps and a label for each speaker, so you can see who said what.
• Saves the finished transcript to your library so it is searchable later.
• Tells you plainly when a recording has more speakers than it could cleanly tell apart, instead of guessing.

The reveal

You drop in a 40 minute interview. A minute later there is a readable transcript, each voice named, every point timestamped. The quote you half remember is right there, ready to copy, and you never had to listen back.

What it doesn’t do

• It does not publish or share your recordings anywhere. Files stay with your assistant.
• It cannot invent labels it is not sure of. On a noisy panel with lots of cross talk, it will tell you which parts are uncertain rather than pretend.

Configuration

• A transcription key: transcription runs through a service (AssemblyAI). You add your key once in Settings, and your assistant handles the rest.
• How you hand it a recording: upload a file or point it at one you have already shared with your assistant.
• Speaker count (optional): if you know how many people are in the recording, tell it, and the labels come out cleaner.

Triggers

Runs on demand: give your assistant a recording and ask for a transcript.