-
Notifications
You must be signed in to change notification settings - Fork 5
Home
Bruno is a speech-to-text-to-AI tool designed to facilitate collaborative learning in groupwork. The prototype application listens to human speech, provides a written transcript that identifies individual speakers, and passes the transcript and a chosen prompt to a LLM.
The backend serves HTML frontend. After the user logs in with Google, they have access to live transcription through audio data passed through the backend to Deepgram (transcription API) and relayed back to the frontend. This transcript is also stored in the database, allowing the user to ask Bruno (GPT with custom prompts and transcript) questions about the conversation.
The frontend is built with html and tailwindcss. The backend is built with the django rest framework, a decision made due to Django's robust feature set, out of the box admin and security capabilities, and compatability with the existing project code written in python. The project also leverages Docker to normalize runtime environments and increase the ease of collaboration.
If you're interested in developing a specialized version of Bruno, we encourage you to fork this repository and implement your changes there. However, if you have ideas or contributions that could broadly enhance Bruno's functionality, please read our contribution guidelines (CONTRIBUTING.md) for more information on how you can contribute to this repository. We greatly appreciate your involvement and input!
v1.0.0 - October 26th, 2023. First open-source stable version.
v1.0.1 - January 6th, 2024. Full replacement of Middlesight with OpenAI
GNU General Public License v3.0 or later
Contributors are able to contribute to the original codebase or create a forked open-source version. We request that you include Principles.md in any derivation of the codebase.
See COPYING to view the full text.