News Research Tool Project

A lightweight and intuitive news research application built with Streamlit and LangChain.
This tool fetches news articles from URLs, processes them via text splitting and embeddings, and stores the results in a FAISS vector database.
Users can then query the content and receive precise, source-backed answers using a RetrievalQAWithSourcesChain.

---

🚀 Features

• Accepts news article URLs for input.
• Loads and extracts content using UnstructuredURLLoader.
• Splits long text via RecursiveCharacterTextSplitter.
• Generates embeddings using OpenAIEmbeddings.
• Stores and indexes embeddings efficiently with FAISS.
• Supports user queries through a RetrievalQAWithSourcesChain, returning answers with source links.

---

⚙️ Installation

• Clone the repository:

  git clone https://github.com/1rishu0/News-Research-Tool-Project.git
  cd News-Research-Tool-Project
  

---

📂 Project Structure

• main.py – Main Streamlit application entry point
• requirements.txt – Python dependencies required for the project
• faiss_store_openai.pkl – Serialized FAISS index for storing embeddings
• .env – Environment variables (not committed; contains your API key)
• LICENSE – Project's open-source license (GPL-3.0)

📸 Demo / Screenshots

---

📜 License

• This project is released under the GPL-3.0 license.
• See LICENSE for details.

---

💡 About

• Compact yet powerful implementation of semantic search over news content
• Combines modern LLM techniques with vector databases
• Builds research tools with accurate answer retrieval using LangChain, FAISS, and OpenAI

---

🙏 Acknowledgments

• Thanks to Codebasics for the helpful tutorial that guided this project.

• Quote from my project:

“I built a News Research Tool with Streamlit and LangChain that fetches news articles from URLs, processes them with text splitting and embeddings, and stores them in a FAISS vector DB. Users can query articles via a RetrievalQA chain to get precise, source-backed insights—showcasing my skills in LLMs and vector search.”