Who it is for
Fee-earners and PAs who run client meetings, hearings follow-ups, or multi-party calls and then spend the next hour reconstructing who said what.
Meeting transcription for legal teams.
Capture a multi-speaker client meeting with speaker labels, edit the transcript, then generate the attendance note instead of starting from a blank page.
The who-said-what
gap.
Generic meeting bots dump text. Legal teams need speaker labels, a chance to correct the record, and a path onto the firm’s attendance-note template.
The hour after the call is spent reconstructing the room.
Speaker identity is guessed later, badly.
The note reaches the firm template only after another retype.
How the job runs
Start meeting mode, correct speakers while it is fresh, then generate the attendance note.
Capture the meeting
Multi-speaker recording with diarisation, microphone, and optional screen capture.
Label the voices
Review speaker labels and correct names while the conversation is still in mind.
Generate the note
Apply the attendance-note template. Approve or reject any Clara rewrite before it is filed.
Who it is for, what it takes,
and what it files.
Audience
Fee-earners and PAs who run client meetings and calls.
Client meetings 
Multi-party calls 
Hearing follow-ups Inputs
Spoken work, with more than one voice in the room.
Speaker diarisation Microphone capture 
Screen capture Outputs & controls
A labelled transcript, then a draft the supervisor will own.
Editable transcript 
Attendance note 
Approval cards 
Workspace access Meetings that leave a usable record
Diarisation separates voices, and the team confirms identity and corrects the transcript. Audio quality, accents, and overlapping speech shape the result, so human verification remains part of the workflow.
Controls
The recording is workspace-scoped. Granular permissions limit who can play it back. Recordings are hosted in the UK by default; transcription is processed by AssemblyAI. SAML SSO is Enterprise-only.
Next step
See attendance notes for the document this workflow produces, or Voice for capture across dictation and meetings together.
Learn more
Meeting transcription questions,
answered.
Can Ardela transcribe meetings with multiple speakers?
Yes. Meeting mode can separate voices into speaker labels. A person should confirm the names and correct the transcript while the conversation is still fresh.
Does speaker diarisation verify who said each sentence?
Diarisation separates voices into speaker labels. The legal team confirms each identity and checks the transcript against the recording.
Can a meeting transcript become an attendance note?
Yes. After the transcript and speaker labels are corrected, Ardela can create a draft attendance note on an available firm template for human review.
Who decides whether a client meeting should be recorded?
The firm does. It should decide whether recording is appropriate, explain the purpose, and obtain any authority required by its professional, contractual, and data-protection obligations.
Leave the meeting
with a transcript.
With the required authority, try Voice on an appropriately protected client call, correct the speakers, and have the supervising fee-earner review the note.