Practice your oral argument with an AI judge.

Your case data stays private

The only things kept on our servers are your email address and your credit balance. There is no table for matters, transcripts, or documents — the case data has nowhere to land.

• The brief file never leaves your browser. Your document is parsed on your own machine. Only the extracted text is sent — in the moment — to the AI provider that generates the bench's questions. We never write it to our database or disk.
• The transcript lives in this tab. Close or refresh and it is gone. Nothing is written to local storage or cookies.
• Audio is transcribed and dropped. No recordings are retained.
• Video never leaves your machine. If you film yourself, the recording is held in your browser. You can save a copy to your own computer from the bench memo page; otherwise it is deleted when you leave. It is never uploaded.
• Business API calls, not consumer chat. Model calls go through commercial API terms, which do not train on the content sent. Providers may hold a request briefly for abuse monitoring before deleting it — that is their retention window, not ours.
• Your usage record is one line. We log that a minute was used and when, so we can bill you correctly. That record contains no case, client, or court information.
• No certifications claimed. We do not assert SOC 2 or anything else that would need an audit. The model provider is a subprocessor; apply your own work-product judgment before uploading a live matter.