Meeting Summarizer

Overview

meeting_summarizer is an automated system designed to handle Slack events, process uploaded files or messages, and interact with Slack users through a bot. The system leverages AWS services including Lambda, API Gateway, SQS, and SSM to create an efficient, serverless processing pipeline.

Architecture

The project integrates various AWS services with Slack to automate file and message processing.

Project Structure

The project is organized into two main directories: infrastructure and lambda_functions.

Infrastructure

Contains the AWS CDK definitions for setting up the infrastructure. It includes:

• api_gateway.py: Sets up the API Gateway.
• lambda_functions.py: Defines the AWS Lambda functions.
• permissions.py: Manages IAM roles and permissions.
• queues.py: Configures the SQS queues.
• ssm_parameters.py: Manages SSM parameters for secure storage of configuration and secrets.

Lambda Functions

Contains the code for Lambda functions and their specific dependencies:

• payload_handler: Quick response Lambda to acknowledge Slack events.

• push_to_sqs_handler: Handles file information retrieval and pushes messages to SQS.

• slack_bot_handler: Processes the Slack events and interacts with Slack API.

  • handlers/slack_handler.py: Main handler for Slack events.
  • services/: Contains service modules like summarizer_services.py, transcription_services.py, and youtube_service.py.
  • utils/: Utility modules like slack_utils.py and ssm_utils.py.

• lambda_layers/ffmpeg_layer: Contains the ffmpeg binaries used for media processing.

Setup and Deployment

1. Prerequisites:

   • AWS CLI installed and configured.
   • Node.js and AWS CDK installed.
   • Slack API tokens and necessary permissions.
   • Setup SSM Parameters using following commands:

     aws ssm put-parameter --name "/meeting_summarizer/openai_api_key" --value <openai_api_key> --type SecureString
     
     aws ssm put-parameter --name "/meeting_summarizer/slack_bot_token" --value <slack_bot_token> --type SecureString
     
     aws ssm put-parameter --name "/meeting_summarizer/input_queue_name" --value <SQSQueueName> --type SecureString
     
     # command to check the parameters
     aws ssm get-parameters-by-path --path "/meeting_summarizer" --query 'Parameters[*].Name'

2. Clone the Repository:

• Clone the repository using following command:

  git clone https://github.com/sou127/meeting-minutes.git
  cd meeting_summarizer

3. Install Dependencies:

• Navigate to each lambda function directory and install dependencies (using a virtual environment is recommended):

  python3 -m venv .venv
  . .venv/bin/activate # on mac/linux
  .\.venv\Scripts\activate # on windows
  pip install -r requirements.txt

4. Deploy with AWS CDK:

• From the root directory, deploy the infrastructure using CDK:

  cdk deploy

5. Configure Slack App:

• Set up event subscriptions and bot permissions in your Slack app configuration.

6. Verify Deployment:

• Test the system by interacting with the Slack bot and checking the logs for AWS Lambda functions.

Usage

After deployment, the Slack bot will be ready to receive events. Users can interact with the bot by mentioning bot along with uploading files or sending youtube url directly to it. The bot will process the received content and respond with the summary of media file.