Skip to content

Private meeting transcription

Private meeting transcription without a meeting bot

Private meeting transcription means controlling where the recording is processed, where the transcript is stored and who can search it. TryScriptorium is designed around a recording you already have: upload it, transcribe it, correct it and keep the result in a searchable workspace.

No bot needs to join the call

For local use, TryScriptorium can record a browser tab or screen with its audio, then turn that recording into text. For an existing Zoom, Teams or Meet recording, upload the file directly. There is no claim that a bot joins or manages your meeting, because it does not.

Make the transcript useful after the meeting

Every line carries a timestamp so you can return to the audio. Speaker labels make a conversation easier to read, although important attributions should always be checked. Search across the library finds a phrase later without reopening every recording. Export the finished version as text, captions, Markdown, JSON or CSV.

Privacy depends on deployment

On this hosted site, a file is sent over HTTPS to the Scriptorium server and is processed there. On your own installation, processing and storage remain on the machine or server you control. Neither mode should be described as the other; see self-hosted transcription for the distinction.

Before you record

Get the consent required where you work, tell participants how the recording will be used and set an appropriate retention policy. Automatic transcription is a working draft, not evidence: verify names, decisions, dates, numbers and quotations against the original audio.

Get started

Try a non-sensitive recording in the hosted workspace, or use the deployment documentation to run the software on your own infrastructure.

Transcribe a meeting recording

Upload a recording you already have and get timestamped, editable, searchable text.

Frequently asked questions

Where is my recording processed?
It depends on the deployment. On the hosted site at tryscriptorium.com, a file is sent over HTTPS to the Scriptorium server and processed there. On your own installation, processing and storage stay on the machine or server you run it on.
Does Scriptorium send recordings to a third-party transcription API?
No. Transcription runs in the application itself using the Whisper family of speech models. A separate, optional AI-recipes feature can send a transcript to a cloud model, but it is off unless an administrator turns it on, and it only runs when a person asks it to.
Does self-hosted Scriptorium work offline?
Transcription does, once the speech model you chose has been downloaded. That download, from Hugging Face, happens once per model size; after that the model is loaded from disk only. The optional speaker-separation model is a one-time download as well.
Does a bot join my meeting?
No. Scriptorium works on a recording you already have, or one you capture from a browser tab on a local installation. It never joins a call.
Is the hosted site suitable for confidential recordings?
Use it for non-sensitive files. If a recording is covered by privilege, medical confidentiality or a contract that forbids third-party processing, self-host instead.
How do I get a self-hosted installation?
Get in touch through the contact page and say what you need to run it for; deployment guidance is provided for local, server-mode and systemd setups.

Keep reading