QuestSecure

Welcome to the QuestSecure repository! 🚀 This assistant is designed to automate the filling of questionnaires efficiently and securely.

📖 Description

QuestSecure simplifies the handling of security questionnaires by:

Automated Filling: Uses a Retrieval-Augmented Generation (RAG) system, built on a knowledge base of internal policies.
Enhanced Security: Ensures the security of data by performing all processing and language model computations on the local hardware.
Ease of Setup: Provides all necessary files and detailed instructions for a quick start.

🛠️ Installation

Before running the application, you need to set up your environment. Here's how you can do it:

Step 1: Clone the Repository

Start by cloning this repository to your local machine:

git clone https://github.com/zaynabreza/DataScienceGroup13.git
cd DataScienceGroup13

Step 2: Install Dependencies

Install all required Python libraries using the provided requirements.txt file:

pip install -r requirements.txt

Step 3: GPU4All and Model Setup

Install GPU4All and download the model you wish to use.

Step 4: Environment Variables

Add the path of the downloaded model, Elastic Search API, path of the Google key, and password of the Elastic Search user to the .env file. Place this file in the src folder:

passwd="YOUR_PASSWORD_HERE"
api_key="YOUR_ELASTIC_SEARCH_API_KEY_HERE"
model_path="YOUR_MODEL_PATH_HERE"
google_key_path="YOUR_GOOGLE_KEY_PATH_HERE"

Step 5: Run the Application

Execute the application using Streamlit:

streamlit run frontend.py

🧠 Building the Knowledge Base

QuestSecure can build a knowledge base from documents for its questionnaire-filling capabilities. Currently, the system supports PDF and Excel files.

Step 1: Organize Your Documents

Place all the PDF and Excel files you want to include in the knowledge base into a single folder.

Step 2: Update Environment Variables

Add the path to your documents folder to the .env file. Set the knowledge_base_path variable with your folder path:

knowledge_base_path="YOUR_PATH_HERE"

Step 3: Generate Embeddings

Run the following command to generate embeddings from your documents, which will be used to power the knowledge retrieval features of QuestSecure:

python generate_embeddings.py

This process may take some time depending on the number and size of the documents.

Support

For support, contact us through the repository issues or pull requests. Contributions to enhance QuestSecure are always welcome! 🌟

Thank you for using QuestSecure!