DICTATION TRANSCRIBER · Controlled audio capture and sequential upload

DICTATION TRANSCRIBER: CONCEPTFlow Native App

The DICTATION TRANSCRIBER demo processes live dictation or uploaded audio inside the native CONCEPTFlow shell, highlights probable PHI and PII, and swaps into a structured results table when processing completes.

Canonical page Audio capture Live dictation Controlled UNMASK review

CONCEPTFlow DICTATION TRANSCRIBER

DICTATION TRANSCRIBER: CONCEPTFlow Native App

This DICTATION TRANSCRIBER app accepts live dictation chunks or uploaded audio files, transcribes them against the local Whisper service, highlights probable PHI and PII, and swaps into a results table when processing completes.

Canonical page Audio upload Live dictation PHI / PII masking

Endpoint: marx/apps/demos/transcriber Backend: Whisper CPU Mode: chunk or file View: CONCEPTFlow native

Input

Mic or audio upload

Record chunked voice dictation or process one or more files sequentially.

Output

Transcription pane

The live transcription view is replaced by a results table once processing is complete.

Masking

PHI / PII tokenization

Flagged spans receive confidence-colored highlighting and anonymous replacement tokens.

Recovery

UNMASK toggle

Review the original unmasked transcript after the output has been anonymized.

Ready

Dictation Transcriber build request. Prompt ready.

Control panel

DICTATION TRANSCRIBER audio capture and sequential upload

Load the bundled sample medical dictation, drop in your own clips, or start a recording session and stream chunks into the local Whisper service.

Audio intake

Drop Audio File Secure local queue

Queued Clip Jones Dictation Demo

Multiple files are processed sequentially in the order selected.

Execution note

Clinical dictation sample for CONCEPTFlow walkthrough. Use the bundled JonesMedical clip to demonstrate live transcription, PHI / PII tokenization, and the UNMASK review flow.

Shown here to keep the walkthrough contextual and auditable.

Queue

Queued clips appear here before they are transcribed.

Dictation Transcriber build request ready.

Accessible output area

Transcribed output and review table

The live transcript panel swaps to the results table when transcription is complete.

Waiting for audio

No transcription yet

Load the sample clip, upload your own audio, or start a recording session to fill this panel.

Transcription and masking audit table
File	Language	Duration	Entities	Status
The audit table is populated once the first transcription completes.