Getting Started
Auto-Process Pipeline

Auto-Process Pipeline

The Auto-Process pipeline takes your audio file and produces a proofread transcript draft automatically. For Legal projects with a DOCX template, it also applies template structure and court formatting — giving you a court-ready draft in one step.

What Auto-Process Does

Auto-Process runs different steps depending on your project type:

General Projects

  1. Transcription — AI converts your audio to text with speaker identification
  2. Proofreading — AI reviews grammar, punctuation, and common transcription errors

Legal Projects (with a template)

  1. Transcription — AI converts your audio to text with speaker identification
  2. Proofreading — AI reviews grammar, punctuation, and common transcription errors
  3. Template injection — Your transcript is placed into your DOCX template at the {{TRANSCRIPT}} placeholder
  4. Section formatting — AI applies court reporting formatting rules section by section

When it finishes, you open the editor and see a draft with all AI changes tracked — ready for your review.

Prerequisites

Project typeRequirements
GeneralAudio file uploaded and transcription complete
LegalAudio file uploaded, transcription complete, and a DOCX template selected
⚠️

Auto-Process for Legal projects requires a DOCX template. If no template is selected, the Auto-Process button will be grayed out. See Uploading Your Own DOCX Template, or choose a built-in template (Standard Deposition or Court Hearing).

Step-by-Step Walkthrough

Create a Legal Project

  1. From the dashboard, click New Project
  2. Select Legal as the project type
  3. Enter a project title (e.g., "Smith v. Jones - Vol. 1")
  4. Fill in optional details: case number, date, witness names

Upload Your Audio

Drag and drop your audio file or click to browse. Supported formats: MP3, WAV, M4A, FLAC, WebM, OGG — up to 4 hours in length.

Attach a Source Document (Optional but Recommended)

If you have a DOCX from your scopist software or a previous draft:

  1. In the project wizard, look for the Source Document step
  2. Upload your DOCX file

The source document is used as context to improve proofreading accuracy — speaker names, legal terminology, and case-specific language are recognized more reliably.

Select a Template

Choose your DOCX template:

  • Standard Deposition — Q/A format with witness examination
  • Court Hearing — Multi-party proceedings
  • Your custom template — If you've uploaded one in Settings

Run Auto-Process

On the project review screen, click Run Auto-Process.

You'll see a progress indicator as each stage completes:

  • Transcribing...
  • Proofreading...
  • Formatting...
  • Done

Processing time for a 2-hour deposition is typically 8–15 minutes.

Review in the Editor

Once Auto-Process completes, click Edit to open the transcript.

All AI changes are shown as Track Changes — green for insertions, red strikethrough for deletions. Review each section and accept or reject changes as needed.

What to Expect in the Editor

After Auto-Process:

  • The transcript is structured with your template sections (cover, appearances, proceedings, certification)
  • Speaker labels follow legal format (MR. SMITH:, THE COURT:, etc.)
  • Q/A indentation is applied to examination sections
  • Examination headers (EXAMINATION, CROSS-EXAMINATION) are in place
  • Placeholders ({{CASE_NAME}}, {{DATE}}, etc.) are filled with your project details
⚠️

Auto-Process produces a starting point, not a finished transcript. AI accuracy is typically 85–95% depending on audio quality. Always review the full document before delivering to clients.

Auto-Process vs Manual Workflow

Auto-ProcessManual
Best forClean audio, straightforward proceedingsDifficult audio, non-standard formatting
Setup time~2 minutes~2 minutes
Processing time8–15 min/2hr audio2–3 min (transcription only)
Output (General)Proofread draft ready to editRaw transcript ready to edit
Output (Legal + template)Fully formatted draftRaw transcript ready to edit
Your reviewAccept/reject AI changesEdit from scratch with audio

Stopping or Re-Running

  • Stop mid-process: Click the Stop button on the progress screen. The pipeline halts after the current stage.
  • Re-run: If the output isn't satisfactory, you can re-run Auto-Process from Project Settings. This creates a fresh transcript and replaces the current one.

Re-running Auto-Process overwrites the current transcript. Export or rename your current version first if you want to keep it.

Troubleshooting

Auto-Process Button is Grayed Out

Check that:

  1. Project type is Legal
  2. An audio file is uploaded
  3. A template is selected in Project Settings

Missing Template Sections

If the cover page or certification is missing in your export:

  1. Confirm your template DOCX has those sections
  2. Check that section headings are recognizable (see Template Section Detection)
  3. The {{TRANSCRIPT}} placeholder must be present in the proceedings section

Formatting Looks Wrong

Auto-Process applies AI formatting which may not be perfect:

  1. Open the editor and use Track Changes to review formatting decisions
  2. Reject incorrect changes and fix manually
  3. Use the AI Format tool on individual sections for targeted corrections

AI Format tool →