Meeting intelligence

Every meeting, capturedtranscribedsummarisedsearchableunderstood

A native desktop app that records system and microphone audio, transcribes it live, and turns the conversation into summaries, decisions, and action items.

Watch releases on GitHub →

See it live

Watch the conversation become intelligence

Recording

00:12:48

Live transcript

Sarah·12:41

Can we lock the launch date before we wrap?

Marcus·12:41

March 14 works if design hands off by Friday.

Priya·12:42

I'll own the migration script and the rollback plan.

Summary

The team aligned on a March 14 launch and assigned owners for the remaining blockers.

Action items

Confirm March 14 launch date with the wider team
Design handoff due Friday — Marcus following up
Priya owns the migration script + rollback plan

What it does

From raw audio to decisions you can act on

One capture pipeline, orchestrated end to end — every external service swappable behind a clean interface.

Native audio capture

Records system and microphone audio together at 16kHz using ScreenCaptureKit on macOS and WASAPI on Windows — no virtual cables.

Live transcription

Streaming speech-to-text turns the conversation into a speaker-labelled transcript as the meeting happens.

AI summaries

A map-reduce pipeline distills the full transcript into a tight summary you can skim in seconds.

Decisions & action items

Owners, decisions, and follow-ups are pulled out automatically, so nothing falls through the cracks.

Semantic search

Find any moment across every meeting by meaning, not just keywords, backed by vector search.

Private by design

Swap the cloud transcriber for an on-prem Whisper model to keep meeting audio inside your own infrastructure.