Skip to main content

Google Meeting Bot

Prerequisites​

Before you begin, ensure the following tools are installed on your system:

Installation Steps​

1. Clone the Repository​

git clone https://github.com/CueMeet/cuemeet-google-bot.git
cd cuemeet-google-bot

Project Folder Structure​

Click to expand folder structure
.
├── Dockerfile
├── Makefile
├── __init__.py
├── app.py
├── buildspec-staging.yml
├── buildspec.yml
├── config
│   ├── __init__.py
│   ├── __pycache__
│   └── settings.py
├── google_meet
│   ├── __init__.py
│   ├── __pycache__
│   ├── bot.py
│   ├── random_mouse.py
│   └── utils.py
├── logger.py
├── monitoring.py
├── out
├── poetry.lock
├── pyproject.toml
├── transcript_extension
│   ├── background.js
│   ├── content.js
│   └── manifest.json
└── utils.py

2. Configure Environment Variables​

Create a .env file and include the following configuration values:

Click to view example .env configuration
DEBUG="True"
HIGHLIGHT_PROJECT_ID="123456"
ENVIRONMENT_NAME="DEV"

3. Configuration Options​

The bot can be customized using command-line arguments:

usage: app.py [-h] [--start-time START_TIME] [--end-time END_TIME] [--min-record-time MIN_RECORD_TIME] [--bot-name BOT_NAME] [--presigned-url-combined PRESIGNED_URL_COMBINED]
              [--presigned-url-audio PRESIGNED_URL_AUDIO] [--max-waiting-time MAX_WAITING_TIME]
              meetlink

Join a Google Meeting to record audio.

Positional Arguments:
  meetlink                        The Google Meet link to join

Optional Arguments:
  -h, --help                      Show help message and exit
  --start-time START_TIME        Meeting start time (JavaScript timestamp in milliseconds)
  --end-time END_TIME            Meeting end time (JavaScript timestamp in milliseconds)
  --min-record-time MIN_RECORD_TIME
                                 Minimum recording time in seconds (default: 7200)
  --bot-name BOT_NAME            Name of the bot in the meeting (default: 'CueMeet Assistant')
  --presigned-url-combined PRESIGNED_URL_COMBINED
                                 Presigned URL to upload a tar file of the recording and transcription
  --presigned-url-audio PRESIGNED_URL_AUDIO
                                 Presigned URL to upload the audio recording
  --max-waiting-time MAX_WAITING_TIME
                                 Maximum waiting time in seconds (default: 1800)

Running with Docker​

Build the Docker Image​

docker build -t google-bot .

View Help Manual​

docker run google-bot /bin/bash -c "python app.py --help"

Run the Bot Locally (Docker)​

docker run google-bot /bin/bash -c "pulseaudio --start && python app.py '<meetlink>' --presigned-url-combined '<url>'"

Local Development with Poetry​

For development without Docker, Poetry is used to manage dependencies and environments.

Setup Instructions​

To set up the project locally using Poetry:

make install

This command will:

  • Install Poetry (if not already present)
  • Configure the virtual environment to reside in the project directory
  • Install project dependencies as defined in pyproject.toml

Activating the Poetry Shell​

To enter the virtual environment:

make poetry-shell

Running the Bot Locally​

Once inside the Poetry shell, you can run:

python app.py --help

Or test a full run with:

python app.py '<meetlink>' --presigned-url-combined '<url>'

Typical Local Workflow​

# Install dependencies and set up the environment
make install

# Enter the Poetry-managed virtual environment
make poetry-shell

# Execute the bot
python app.py --help

Output Artifacts​

  • Audio File: Captured in .opus format.
  • Transcription: Saved as a .json file.
  • Combined Archive: Both files are compressed into a .tar archive and uploaded to S3 using the provided --presigned-url-combined.

Note: Output of each of these is stored in the /out folder of the parent directory.

Uploading Process​

  • The audio file is uploaded separately using the --presigned-url-audio.
  • A tar archive (.tar) containing the audio and transcription is uploaded using the --presigned-url-combined.