Plaud Multi-Note Sync

I wear a Plaud AI recorder clipped to my shirt. It captures meetings, therapy sessions, walks where I think out loud. Plaud’s app gives you transcripts and AI-generated summaries, but its export only gives you the transcript — no notes, no action items, no outlines.

I needed all of it. The notes are where the pattern recognition happens. When three different colleagues mention the same friction point across three separate meetings, that’s a training gap. When my therapist and I circle back to the same trigger across four sessions, that’s a pattern worth tracking. The transcript alone doesn’t surface those.

So I reverse-engineered Plaud’s API. Their mobile app makes authenticated requests to pull session data — I intercepted those, figured out the auth flow (OAuth token refresh), and wrote a sync script that pulls every session with all its metadata: transcript, AI summary, action items, notes, outlines. 101 sessions synced so far.

The real value isn’t the sync itself — it’s what happens after. Each session becomes a JSON file in my project. Claude can read them. When I’m building a training deck about communication best practices, I tell Claude to check my Plaud sessions first — the raw transcripts contain more actionable detail than any formal document. A 30-slide deck on report formatting came directly from notes I took during a colleague’s feedback session.

The sessions are also where skills come from. I watched my boss validate a deliverable six times across six meetings. Each time she did the same thing: divided two numbers, checked if the ratio matched her expectation. I wrote that pattern into a skill file. Now every pipeline I build pre-passes her check before she sees it.