Practice a spoken mock trial against AI witnesses and opposing counsel on a fictional case.
Upload your own case packet and have it compiled into a structured case for a hearing.
Review a transcript-grounded coaching report after conducting a hearing.
| niksh-hiremath/suits | 0xradioac7iv/tempfs | 7vignesh/pgpulse | |
|---|---|---|---|
| Stars | 0 | 0 | 0 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | hard | moderate | moderate |
| Complexity | 4/5 | 3/5 | 4/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Needs Node.js, a Python toolchain, a Convex account, an OpenAI API key, and multiple generated secrets kept in sync.
SUITS is a courtroom simulator built for education, not for real legal work. It lets a user pick one of a few pre made fictional cases, or upload their own case documents, and then take part in a spoken mock trial against AI characters playing other courtroom roles, such as witnesses or opposing counsel. After the hearing, the user gets a coaching report grounded in the actual transcript of what was said. The project is explicit that it does not give legal advice or predict how a real case would turn out. The application is a Next.js website backed by Convex for storing cases and trial data, and it uses an AI model for compiling case details ahead of time and for reasoning through each courtroom role during the hearing itself. Speech is handled by a separate local Python program running on the user's own machine, which converts spoken words to text and generates spoken responses back, so the whole hearing can happen by voice with no text box needed during a real session. The courtroom itself is rendered as an animated 3D scene in the browser. Every hearing keeps a full, privacy conscious record tied to the person who ran it: the sequence of events, the transcript, what happened procedurally, information about which AI calls were made, and a metadata only log of audio activity, all of which can be exported as JSON. Setting it up requires Node.js, a Python toolchain, and a Convex account, along with an API key for the AI model. Several secret values also need to be generated and kept in sync between the web app and Convex. A real microphone based hearing over a network connection and a finished production deployment are both still works in progress rather than finished features.
A voice driven mock courtroom simulator for educational trials, with AI witnesses and a transcript based coaching report.
Mainly TypeScript. The stack also includes Next.js, TypeScript, Convex.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.