Keeper

Songwriters record long voice memos: 20 minutes of noodling, talking to themselves, and singing the same chorus six different ways. The good take is in there somewhere, and finding it means scrubbing the whole file by hand. Most memos never get opened again. Keeper finds the takes for you. Upload a memo and it detects where you are singing or playing versus talking, cuts each take, and listens to what you said around it ("that's the chorus", "keep that one", "scrap that"). Then it files each take into a song canvas organized by section. You can play takes back, rearrange them, and export the arrangement as a WAV or every take as stems. It is built for the way songwriters actually work. You don't have to tag anything while recording: you talk the way you already do, and Keeper does the filing. Live: https://mhashamhussain2001--keeper-web.us-east.modal.run