whisperleafai

whisperleafai was built for the moment after a lecture, a meeting, or an interview, when the conversation's over and you're left with forty minutes of audio and no clear way to learn from it. Open the app, hit record, and by the time you stop you already have a clean transcript, a structured summary, and an optional quiz generated from the content.

It runs everywhere Flutter does (iOS, Android, web, Windows, macOS, Linux) with offline support where it matters and a Word-template export for the times you need the output in a document, not an app.

The build was an experiment in how fast a single developer could go with AI coding agents doing the unglamorous work. The answer turned out to be: faster than we expected. whisperleafai is proof-of-method as much as product.

We're finishing the transcription pipeline (multi-speaker diarization, better noise handling) and tuning the summary quality so study guides are actually useful, not just plausible-looking.

Planned: shared notebooks, a light spaced-repetition layer for the quizzes, and a desktop capture mode for recording calls and online lectures.

No public build yet. If you're a student, a researcher, or anyone who lives in meetings, we'd love to hear what you'd want from a tool like this.