Architectural Blueprints for RAG Automation: Advanced Document Understanding using Vertex AI Search

This repository offers a comprehensive toolkit for building and evaluating RAG pipelines specifically designed for question answering on documents. Leveraging Vertex AI Search and Gemini Pro, it streamlines document ingestion, enhances search accuracy, and enables robust evaluation of retrieval and answer quality. Explore the provided code and detailed [Medium article] to learn how to create and evaluate RAG pipelines that deliver accurate and reliable responses to your inquiries.

Setup Instructions

Follow these steps to set up the RAG Automation project environment on your local machine.

1. Clone the Repository

First, ensure that you have git installed on your machine and then clone the repository using:

git clone <repository-url>
cd <repository-directory>

Replace <repository-url> and <repository-directory> with the actual URL and directory name of your repository.

2. Create a Virtual Environment

Create and activate a virtual environment to manage the project's dependencies:

python3 -m venv .vais-rag-patterns
source .vais-rag-patterns/bin/activate

3. Install Dependencies

Upgrade pip to the latest version and install required packages from requirements.txt:

python3 -m pip install --upgrade pip
pip install -r requirements.txt

4. Set Environment Variables

Append the project directory to your PYTHONPATH environment variable:

export PYTHONPATH=$PYTHONPATH:.

5. Configuration Files

Make sure to have your credentials and config.yml files set up in the project directory as these files are essential for the correct functioning of the project.