Somebody Make This: An audio transcription mobile app. Where you can record & export your voice notes to any note taking destination.
I've been looking for this for ages and can't seem to find one.
I want it integrated with Android intents/Google Assistant or Siri Intents so that you could go
"Hey Google/Siri, start Transcribing in Xnote"
It would open a recorder that lets you speak your notes at your own leisure and it transcribes them into a text note.
Then you could configure it to send to a bunch of destinations. Like Notion, Google Keep, Obsidian, Excel, Sheets etc automatically.
-
The record function needs to be non-timed, like google assistant recordings listen for silence before terminating the recording which doesn't work. It needs to be open until you interact with the screen & press Stop.
-
It would be cool to also be able to 'snippet the last 30 seconds of audio' in case you were listening to a podcast but not essential as MVP.
-
You could add an AI layer on this where it would optionally summarise your voice notes or clean them up. You could use a more enhanced voice transcription API like openAI's WhisperAi to reduce errors too.
-
The target markets could be professors wanting to note something down. People whilst driving taking things to check on later. Whilst listening to a podcast being able to snip awesome quotes or anecdotes. A hands free way to take notes without having to think about app and transcription etc.
What do you think?
The open-ended recorder is the better pick. Time-boxed notes always feel a bit fake to me, because the thought usually shows up after the timer would’ve ended. If the goal is just to get something down before it slips away, DictaFlow is in that same lane on the live-capture side, just without the extra export step.
I built something close to this — Cider Press (available on GitHub, i can't post links yet says indiehackers). It works with your existing Apple Voice Memos (so you keep using the recorder you already have), copies them into a local database, and transcribes them in bulk using Whisper with Metal GPU acceleration. You can then export audio or transcripts anywhere, and there's a built-in integration with Google NotebookLM. The key insight for me was: don't replace Voice Memos — liberate them. People already have a perfectly good recorder in their pocket. What's missing is the pipeline from recording to searchable, exportable, AI-queryable text. Free and open source.