This project demonstrates the use of Pinecone for creating a vector embedding database and Langchain for handling language operations with OpenAI's ChatModel. The goal is to create a question answering chatbot from PDFs, create an index, and query the data efficiently.

1. Clone the repository:

   git clone https://github.com/VirajDeshwal/mdc-RAG
   cd mdc-RAG

2. Install Anaconda:

   • Download and install Anaconda from here.

3. Create and activate a new conda environment:

   conda create --name rag python=3.10
   conda activate rag

4. Install the required dependencies:

   pip install -r requirements.txt

5. Configure environment variables:

   • Create a .env file in the root directory.
   • Add your OpenAI and Pinecone API keys:

     OPENAI_API_KEY=your_openai_api_key
     PINECONE_API_KEY=your_pinecone_api_key

6. Add your PDF files:

   • Place the PDF files you want to process in the input_src folder.

7. Create the vector database and index the data:

   python run.py

8. Run a query against the indexed data:

   python pinecone_query.py

• run.py: Script to create the vector index and store the data.
• pinecone_query.py: Script to query the Pinecone vector database.
• requirements.txt: List of dependencies required for the project.
• input_src/: Directory to place the PDF files to be processed.
• utils/: Utility functions and modules used in the project.
• .env: Environment file to store API keys (not included in the repository).

1. Indexing Data:

   • Ensure your PDFs are in the input_src folder.
   • Run python run.py to create the vector index.

2. Querying Data:

   • Run python pinecone_query.py to perform queries on the indexed data.

• Make sure to replace your_openai_api_key and your_pinecone_api_key with your actual API keys in the .env file.