textract-bedrock-document-insights

Low Latency Document Analysis with AWS

A sample Streamlit application for analyzing documents using AWS Textract and Bedrock - Amazon Nova Micro Model. Users can upload documents (images or PDFs) and get quick insights.

Architecture

Features

• Document upload (supports PNG, JPG, JPEG, PDF)
• Text extraction using AWS Textract
• Document analysis using AWS Bedrock (Nova Micro model)
• Custom prompts for tailored analysis
• Execution Time Tracking

Installation

1. Clone the repository:

   git clone https://github.com/aws-samples/textract-bedrock-document-insights.git
   cd textract-bedrock-document-insights
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

Configuration

Set the following environment variables:

• S3_BUCKET: The name of your S3 bucket for document storage
• AWS_REGION: The AWS region to use (defaults to "us-east-1" if not set)

You can set these variables in your environment or optionally use a .env file as below:

1. Create a .env file in the root directory of the project:

   touch .env
   

2. Add the above environment variables to the .env file
3. The application will automatically load these variables using python-dotenv.

Note: Make sure to add .env to your .gitignore file to keep these details secure.

Usage

Run the Streamlit application:

streamlit run src/main.py

Then, open your web browser and navigate to the URL provided by Streamlit.

Dependencies

• boto3
• streamlit
• python-dotenv

Make sure you have the necessary AWS credentials configured to access S3, Textract, and Bedrock services, along with model access to Amazon Nova Micro.

Optional: Running with Virtual Environment

If you prefer to run this project in a Python virtual environment, follow these steps:

1. Create a virtual environment in your project directory:

   python -m venv venv

2. Activate the virtual environment:

   • On macOS/Linux:

     source venv/bin/activate

   • On Windows:

     venv\Scripts\activate

3. Install the required packages:

   pip install -r requirements.txt

4. Run the application:

   streamlit run src/main.py

5. When you're done, you can deactivate the virtual environment:

   deactivate

Note: You'll need to activate the virtual environment each time you want to run the project.