stephenc222/example-retail-feedback-analysis-tool

Example project to explain a practical way to use a multimodal LLM for a business purpose.

Retail Feedback Analysis

Setup

1. Clone the repository:

   git clone git@github.com:stephenc222/example-retail-feedback-analysis-tool.git
   cd example-retail-feedback-analysis-tool

2. Create a virtual environment:

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

3. Install dependencies:

   pip install -r requirements.txt

4. Set up environment variables:

   • Create a .env file in the root directory of the project.
   • Copy the contents of .env.example into .env and fill in your Google API key.
   • Need to setup a Google GenAI billing account

5. Generate a report:

   sh generate_report.sh

6. Build the client:

   sh build_client.sh

7. Start the Flask app:

   sh start_server.sh

After the Flask app is running, you can access the client at http://localhost:5000.

Dependencies

• Python 3.8+
• google-generativeai
• dotenv
• ffmpeg-python
• librosa
• noisereduce
• numpy
• soundfile
• pydub
• spacy
• opencv-python
• Pillow

Additional Setup

• Install ffmpeg if not already installed:

  pip install ffmpeg-python

Error Handling

Custom exceptions are used throughout the codebase to handle specific errors in data ingestion, preprocessing, analysis, integration, and reporting steps.