GitHub

Meetily Local-First Meeting Privacy

Trace data paths, owners, and external edges

7 chapters75 min total
Keywords
Meetilylocal-firstprivacytranscriptionsummariesWhisperParakeetSQLitetelemetryOllama
Overview

Meetily’s current Tauri desktop application can keep capture, Whisper or Parakeet transcription, storage, and selected summaries on the user’s infrastructure, but local-first is configurable rather than absolute. This series traces audio and speech segmentation, model and hardware costs, SQLite and file artifacts, summary-provider boundaries, telemetry, updates, and recovery. It shows what local control buys—visibility, optionality, and reduced routine vendor custody—while preserving limits around permissions, platform behavior, model provenance, encryption, deletion, provider policies, and bounded failures.

Learning outcomes
  • Meetily’s core capture, local transcription, and storage paths can run on the user’s infrastructure.
  • Local transcription does not make summary generation local: transcript text may go to a cloud or custom endpoint.
  • Local storage provides inspectable files, recovery, and retranscription, but not proven encryption or complete deletion.
  • Privacy depends on permissions, models, platform paths, telemetry, updates, and the operator’s retention choices.
Keywords
Meetilylocal-firstprivacytranscriptionsummariesWhisperParakeetSQLitetelemetryOllama

Chapters

Preview every chapter · Unlock full text, audio, and video from chapter 4

Plans
Ch.1-3 FreeCh.4-7 Premium

FAQ

What does local-first mean in Meetily?

In Meetily, local-first means the supported desktop app can keep capture, local transcription, storage, and selected summary routes on the user’s infrastructure, not that every network edge disappears.

Does Meetily send meeting audio or transcripts to the cloud?

Meetily’s supported local recording path need not send audio to a Meetily server, but a chosen cloud or remote custom summary endpoint can receive transcript text and context; model downloads, updates, and opt-in analytics are separate external paths.

Is Meetily’s local storage encrypted and fully deleted?

The examined code does not establish application-level encryption or complete deletion of every file and backup. Database rows are deleted transactionally, but meeting folders and copies require separate inspection.

Which Meetily summary option is most private?

No universal option is most private: local Ollama or built-in AI keeps inference on local infrastructure, while custom and cloud routes depend on the actual endpoint, operator, and provider controls.