rupamsamanta123/PDF-QA-RAG-System
📄 Transform your PDF documents into actionable insights with this RAG-based Question-Answering App for efficient and accurate responses.
What's novel
📄 Transform your PDF documents into actionable insights with this RAG-based Question-Answering App for efficient and accurate responses.
Code Analysis
3 files read · 2 roundsA single-file Streamlit app that extracts text from one uploaded PDF, chunks it, stores it in an in-memory Chroma vector store using Ollama embeddings, and answers a single query by stuffing retrieved chunks into a prompt sent to llama3.1 via Ollama.
Strengths
The code is concise, readable, and correctly wires together standard LangChain/Ollama/Chroma components into a working RAG pipeline. Function names are clear and the flow is easy to follow.
Weaknesses
No error handling whatsoever (no try/except, no validation of Ollama availability, no handling of empty/corrupt PDFs), zero tests, no version pinning, and the README falsely claims multi-PDF upload and describes a desktop installer that doesn't exist. The implementation is a thin wrapper around well-known library calls with no custom logic, no persistence, no conversation memory, and no re-ranking.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
7
Contributors
2
Files
6
Active weeks
3
Repository
Language
Python
Stars
1
Forks
0
License
MIT