Papers QA Agent

A LangChain-powered question-answering system for scientific papers using Ollama LLM and ChromaDB for vector storage.

Features

• Load and process PDF papers automatically
• Smart document chunking and embedding generation
• Interactive Q&A command line interface
• Vector similarity search using ChromaDB
• Progress tracking for document processing

Prerequisites

• Python 3.8+
• Ollama running locally
• PDF papers to analyze

Installation

1. Clone the repository:

git clone <repository-url>
cd papers-qa-agent

2. Setup and dependencies :

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install requirements
pip install -r requirements.txt

4. Configure environment variables in .env:

PROJECT_NAME=scientific_papers_qa
PAPERS_DIR=./papers
OLLAMA_MODEL=deepseek-r1:14b 
OLLAMA_EMBEDDING_MODEL=snowflake-arctic-embed:335m-l-fp16
CHROMA_DB_DIR=./chroma_db
OLLAMA_BASE_URL=http://localhost:11434

Usage

1. Add PDF papers to the papers directory
2. Run the application:

python __main__.py

Available Commands

• audio: Use the audio to text STT feature
• help: Show available commands
• metadata: Display information about loaded papers
• exit: Quit the application

Additional notes:

pyaudio might require additional system-level dependencies:

• On Windows: No additional requirements
• On Linux: sudo apt-get install python3-pyaudio
• On macOS: brew install portaudio