In My Head 0.2 Release Notes

Create a Session from selected recordings, add a recording to an existing Session, or start a recording directly from a Session.

Choose whether to add a recording to a Session as soon as recording stops; transcription can continue while that choice is made.

Read and refine one merged Session transcript, with its recording count and a direct route back to each source recording.

Browse and sort Sessions independently by name or date.

Standalone recordings have their own transcript version history, with edit, copy, and Model polish actions.

Sessions have an independent merged transcript and version history. Once a recording is part of a Session, its individual transcript is read-only and links back to that Session instead.

Resume a standalone recording to transcribe only newly captured audio and append it to the existing transcript; this creates a reviewable version rather than rewriting earlier edits.

Enable automatic Model polish for a recording’s first transcription, or run it manually when you want it. Each polish is an explicit transcript version.

Recordings can still be added to a Session when the model is unavailable or automatic polish is turned off.

Generate concise titles for recordings and Sessions on-device after their transcripts are ready.