§Court Transcript Platform

How it works

The platform turns a finished transcription into a filing-ready, jurisdiction-compliant court transcript. The workflow is five steps, all in one workspace:

  1. 1. Import a transcript

    Upload the finished transcript PDF from your transcription provider (e.g. a Rev.com export). The importer detects the layout, parses every speaker turn and word, and reports exactly what it found before anything is saved.

    • Every word is stored individually, so later edits are tracked and undoable.
    • Unknown layouts are rejected with a clear reason instead of a bad parse.
    • You confirm the parse summary (turns, speakers, word count) before the transcript is created.

  2. 2. Map speakers and case details

    In the workspace, give each detected voice its court identity — name and role. Roles drive the formatting: a Judge renders as THE COURT, attorney/witness exchanges render in Q/A format automatically.

    • Merge duplicate voices (diarization often splits one person into several labels) — merges are reversible.
    • Unidentified voices like “Speaker 4” stay visible until you map them; export is blocked until mapping is complete.
    • Fill the case information panel: court, case number, case name, judge, hearing date, and the certifying reporter. The header, caption, and certification pages are built from these fields.

  3. 3. Pick the court format

    Choose the destination court in the Court format selector. The document reformats instantly to that jurisdiction's published rules — grid, line numbers, margins, Q/A indents, caption, certification wording, and required indexes.

    • If the court field matches a known jurisdiction, the matching format is suggested automatically.
    • Formatting is derived, never stored: switching formats never touches the words.
    • Available today: California Superior (audio transcription), U.S. District Court, Florida (Rule 2.535), Texas (Uniform Format Manual), and New York (Part 108).

  4. 4. Review and correct

    The center pane is the formatted document exactly as it will file — numbered lines, page furniture, frame. Click any word to correct it; use find & replace for recurring errors; name-consistency suggestions catch misspelled party names.

    • Undo/redo covers every edit (Ctrl+Z / Ctrl+Shift+Z).
    • Edits made offline are queued locally and replayed when you reconnect.
    • Reformatting after an edit is incremental — even 500-page transcripts stay responsive.

  5. 5. Approve and export

    The QA gate lists everything that still blocks filing — unmapped speakers, missing metadata, a missing certifying reporter. When every check passes, approve the transcript and export PDF or DOCX.

    • Exports carry the full apparatus the format requires: caption page, certification page, word index or witness/exhibit indexes.
    • The certification page is generated, but the attestation belongs to the human reporter — the software never signs.
    • Every export is recorded with the revision, format, and approving reviewer that produced it, and downloads are access-controlled.

What “compliant” means here

Each court format is a measured, tested ruleset — built from the jurisdiction's published standards (e.g. California's Minimum Transcript Format Standards and Rules of Court 2.100–2.111) and, where available, from certified transcripts the court actually accepted. Layout is verified against those references by automated tests on every change.

Import a transcript