All features

Transcribe recordings

Upload a consultation, a phone call or a voice note to the matter, and get a transcript with its speakers told apart, named as the matter's people, that plays from any line and that Squire™ reads like any document.

What a client said in a consultation, or the other side said on a call, is evidence you will want to find again. Upload the recording to the matter and Squire™ transcribes it, tells the speakers apart and lets you say who each one is. The transcript becomes the file's text: searchable, quotable and cited like any document - with the recording a click away from every line.

A recorded call on the matter finishes transcribing; on its page Suggest who is who proposes the matter's people for Speaker 1 and Speaker 2, the lawyer accepts both and the transcript reads with their names, and clicking a speaker's line at 00:03:41 plays the recording from 3:41.

How it works

  1. 1

    Upload the recording

    Add it to the matter's documents as you would a PDF - MP3, M4A, WAV, OGG, Opus, FLAC, AAC or WebM audio, up to 200 MB and four hours. It is transcribed in the background; the file shows Preparing audio and then Transcribing while you carry on.

  2. 2

    Listen and check

    On the file's page the transcript sits under the recording's player, a time on every turn. Click a time and the recording plays from there, so a figure, a date or a "not" is checked against what was actually said.

  3. 3

    Say who is speaking

    Each speaker comes with two of their lines to play. Type who they are: one of the matter's people, someone else in the firm's records, a new party, or just a name. Suggest who is who proposes a person for each speaker from introductions and names used in the conversation, for you to accept, and Same person as… joins two speakers that are one voice.

  4. 4

    Ask about it

    The transcript is the file's text, with the speakers' names in it. It is summarised and indexed like a document, so in the matter's chat you can ask what the client said about the deposit and get the answer with the time it was said, cited to the file.

Checked against the recording

  • Every turn has its time, and plays from it
  • The transcript says it was made automatically and may contain errors
  • Renaming a speaker rewrites the text everywhere Squire reads it

The client's voice stays the client's

  • No one is identified by their voice - a member of your firm says who is speaking
  • Transcribed by a provider Squire chooses for the region, never one a firm can change
  • The recording and transcript are encrypted under your firm's key

English only for now, and audio only - not video. A poor recording or people talking over each other will leave mistakes, so check a passage before you rely on it.

Speaking rather than typing? Dictate into a chat, a note, a todo or a comment.

More in the FAQ: transcribing recordings and who hears the audio.

Open a matter