Audio Recording & Private Meeting Memory

Phasoric turns meeting audio and voice recordings into structured, source-backed meeting records. Local transcription keeps audio on the device; connected processing is a separate opt-in route whose provider terms apply.

Explicit Recording Consent & Safety

Before capturing audio:

Phasoric displays a mandatory Recording Consent Acknowledgement overlay.
Audio capture cannot start without an explicit user interaction and affirmative consent confirmation.
Recording indicators remain visible throughout the active session.

Recording & Importing Meetings

1. Open Meeting Memory from the side navigation or click the Record Audio microphone icon in any note.

2. Select your microphone or drop an existing .mp3, .wav, or .m4a file into the upload zone.

3. Choose your transcription processing mode:

- Local Processing (Default): Audio is transcribed on your device using WebAssembly / local Whisper models without sending audio over the network.

- Connected Processing (Opt-in): Uses cloud transcription models for extended multi-hour meetings when explicit network consent is enabled.

Speaker & Timestamp Intelligence

Speaker Labels: Phasoric recognizes explicit Name: speech or Speaker 1: speech labels already present in a transcript. Automatic diarization is not claimed.
Transcript Jump Links: When source lines include timestamps, reviewed items link back to that time; otherwise they link to the source line.

Automated Extraction & Action Workflows

After transcription completes, Phasoric automatically analyzes the conversation to propose structured candidates:

Decision Candidates: Converts agreed decisions into Strata 2.0 decision proposals.
Commitments & Due Dates: Identifies assigned follow-ups and adds them as Pulse 2.0 signals.
Forge Action Drafts: Generates task lists, meeting summary notes, or follow-up email drafts for one-click review and approval.
Source Provenance: Reviewed items retain a link to the available transcript timestamp or source line. Rejecting a candidate does not mutate the underlying transcript or audio file.