0x77dev/meetingscribe

MeetingScribe is an AI-driven command-line tool designed to streamline your meeting experience by handling transcription, translation, and note-taking. Effortlessly generate accurate translation/transcription in English from audio file. Additionally, the tool intelligently creates meeting notes, summaries, and identifies action items.

MeetingScribe

MeetingScribe is an AI-driven command-line tool designed to streamline your meeting experience by handling transcription, translation, and note-taking. Effortlessly generate accurate translation/transcription in English from audio file. Additionally, the tool intelligently creates meeting notes, summaries, and identifies action items.

Prerequisites:

1. Specify OpenAI API Key:

export OPENAI_API_KEY=<your-openai-api-key>

2. Install FFmpeg

Installation:

using pip

pip install meetingscribe

using docker

export OPENAI_API_KEY=<your-openai-api-key>

docker run -it -e OPENAI_API_KEY=$OPENAI_API_KEY ghcr.io/0x77dev/meetingscribe --help

Usage:

meeting [OPTIONS] COMMAND [ARGS]...

Options:

• --install-completion: Install completion for the current shell.
• --show-completion: Show completion for the current shell, to copy it or customize the installation.
• --help: Show this message and exit.

Commands:

• interactive: Chat for answering questions based on the...
• process: Transcribe (and optionally translate to English) audio file into SRT file
• srt2txt: Transform SRT file to TXT file
• summarize: Generate meeting summary, notes, and...

meeting interactive

Chat for answering questions based on the provided SRT file

Usage:

meeting interactive [OPTIONS]

Options:

• --input-srt-file TEXT: [default: output.srt]
• --model TEXT: [default: gpt-3.5-turbo]
• --temperature FLOAT: [default: 0.5]
• --help: Show this message and exit.

meeting process

Transcribe (and optionally translate) audio file into SRT file Translation will translate from source language to English

Usage:

meeting process [OPTIONS] INPUT_AUDIO_FILE

Arguments:

• INPUT_AUDIO_FILE: [required]

Options:

• --output-srt-file TEXT: [default: output.srt]
• --source-language TEXT
• --segment-length INTEGER: [default: 600000]
• -y, --yes: Skip confirmation prompt
• --model TEXT: [default: whisper-1]
• --help: Show this message and exit.

meeting srt2txt

Transform SRT file to TXT file

Usage:

meeting srt2txt [OPTIONS]

Options:

• --srt-file TEXT: [default: output.srt]
• --output-file TEXT: [default: output.txt]
• --help: Show this message and exit.

meeting summarize

Generate meeting summary, notes, and action items from SRT file

Usage:

meeting summarize [OPTIONS]

Options:

• --input-srt-file TEXT: [default: output.srt]
• --output-summary-file TEXT: [default: output.md]
• -y, --yes: Skip confirmation prompt
• --model TEXT: [default: gpt-3.5-turbo]
• --help: Show this message and exit.

meetingscribe's Issues

Offline-first features

• #9
• #10

Move to LiteLLM to support Ollama and other LLMs

Give the ability to use whisper-cpp instead of OpenAI's cloud