Serverless-observability

Demo - MusicQA: AI-Powered Music Question Answering System

MusicQA is an AI-powered system that answers questions about music using AWS services and the CDK (Cloud Development Kit) for infrastructure deployment.

This project implements a serverless architecture to process user queries about music, leveraging natural language processing and a knowledge base to provide accurate and informative responses. The system is designed to scale automatically and handle a wide range of music-related questions efficiently.

Repository Structure

The repository is organized as follows:

• app.py: The main entry point for the CDK application.
• cdk.json: Configuration file for the CDK project.
• requirements.txt: Python dependencies for the project.
• musicqa/: Directory containing the main application code.
  • musicqa_stack.py: Defines the AWS resources using CDK constructs.
• aws_xray_sdk/: Directory containing the AWS X-Ray SDK for Python.
• botocore/: Directory containing the Botocore library, which is a low-level interface to AWS services.
• bin/: Directory containing utility scripts.
  • jp.py: A command-line interface for the JMESPath library.
• source.bat: Windows batch script for activating the Python virtual environment.

Usage Instructions

Installation

1. Ensure you have Python 3.7 or later installed.
2. Install the AWS CDK CLI:

   npm install -g aws-cdk
   

3. Clone this repository:

   git clone https://github.com/your-repo/musicqa.git
   cd musicqa
   

4. Create and activate a virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # On Windows, use `.venv\Scripts\activate`
   

5. Install the required dependencies:

   pip install -r requirements.txt
   

Deployment

1. Configure your AWS credentials:

   aws configure
   

2. Synthesize the CloudFormation template:

   cdk synth
   

3. Deploy the stack:

   cdk deploy
   

Using the MusicQA System

After deployment, you can interact with the MusicQA system through the provided API Gateway endpoint. Send POST requests to the endpoint with your music-related questions in the request body.

Example:

curl -X POST https://your-api-gateway-url/prod/ask -d '{"question": "Who wrote Bohemian Rhapsody?"}'

Data Flow

1. User submits a question through the API Gateway.
2. The request is routed to a Lambda function.
3. The Lambda function processes the question using natural language processing.
4. The processed question is used to query a knowledge base (e.g., DynamoDB or Amazon Kendra).
5. The retrieved information is formatted into a response.
6. The response is sent back through the API Gateway to the user.

[User] -> [API Gateway] -> [Lambda] -> [NLP Processing]
                                   -> [Knowledge Base Query]
                                   -> [Response Formatting]
         [API Gateway] <- [Lambda] <- [Final Response]
[User] <-

Infrastructure

The project uses AWS CDK to define and deploy the following resources:

• Lambda: Processes questions and generates responses.
• API Gateway: Provides the HTTP endpoint for user interactions.
• DynamoDB: Stores the music knowledge base (optional, depending on implementation).
• Amazon Kendra: Provides intelligent search capabilities (optional, depending on implementation).
• CloudWatch: Monitors and logs system activities.
• IAM: Manages permissions for the various components.

The exact resources and their configurations are defined in the musicqa_stack.py file.