📄 RAG-based-on-PDF

RAG-based-on-PDF is a Retrieval-Augmented Generation (RAG) system that lets users ask questions and get document-grounded answers. In this version, the system is built on a single scraped web page from VS Code documentation.

It uses FAISS vector search, HuggingFace embeddings, and OpenRouter LLM via a Python backend, with a Streamlit frontend for interactive usage.

---

🚀 Key Features

• Ask natural language questions from web-scraped documents
• Efficient retrieval using FAISS embeddings
• Context-aware answers using OpenRouter LLM
• Streamlit frontend for interactive Q&A
• Fully Python-based and modular architecture

---

🧠 Technology Stack

• Python: FAISS, LangChain, HuggingFace Transformers, Streamlit
• OpenRouter LLM (Cloud-based)
• Pickle for embeddings storage

---

📂 Project Structure

RAG-based-on-PDF/
├─ .venv/                      # Python virtual environment
├─ scraped_docs/               # Scraped text files
│   └─ getting_started.txt      # VS Code Getting Started page
├─ .env                        # Environment variables
├─ app.py                       # (Optional backend for future extensions)
├─ App_streamlit.py            # Streamlit frontend for Q&A
├─ faiss_index.index           # FAISS index
├─ faiss_index.pkl             # FAISS metadata
├─ rag_engine.py               # RAG engine implementation
├─ web_scraping.py             # Script to scrape VS Code page
├─ README.md
├─ requirements.txt
└─ other project files

---

⚙️ Setup Instructions

1️⃣ Environment Setup

python -m venv .venv
.venv\Scripts\activate      # Windows
# source .venv/bin/activate  # Linux/Mac
pip install -r requirements.txt

Add your OpenRouter API key to .env:

OPENROUTER_API_KEY=your_openrouter_api_key_here

---

2️⃣ Run Streamlit Frontend

streamlit run App_streamlit.py

Open in browser: http://localhost:8501

---

▶️ Usage

1. Open the Streamlit app in your browser
2. Type a question in the input box
3. The system retrieves relevant content from the scraped VS Code page and generates an answer using OpenRouter LLM

Optional: You can show the retrieved context to see which text chunks were used.

---

🎥 Demo

Watch the demo video here: 👉 https://drive.google.com/drive/folders/1hQ1H5QJlVAmvm9sM1MyERI-o9GXKp5-z?usp=drive_link

🧩 RAG Workflow Overview

1. The VS Code page is scraped and saved as text (getting_started.txt)
2. The text is split into smaller chunks
3. Chunks are embedded using HuggingFace embeddings and stored in FAISS
4. User question is embedded
5. FAISS retrieves the most relevant chunks
6. Retrieved chunks are passed to OpenRouter LLM
7. LLM generates a factual, document-grounded answer

---

📌 Important Notes

• Answers are strictly based on the scraped page content
• If no relevant context is found, it returns “Not found in document”
• No additional internet or external APIs are required during inference (except for OpenRouter LLM calls)

---

👩‍💻 Author

Iqra Khan AI Engineer | RAG Systems | LLM Applications | Streamlit

---

📄 License

For educational and demonstration purposes only.

---