🧠 LangChain QA System with Chroma & FAISS

A document-based Question-Answering system using LangChain, HuggingFace Embeddings, Ollama (Mistral), Chroma, and FAISS — all running locally.

---

📄 Overview

This project enables you to ask natural language questions over a set of banking and payment-related documents. It uses:

• 🧩 LangChain to orchestrate the chain
• 🧠 HuggingFace Embeddings (all-MiniLM-L6-v2)
• 🗃️ Chroma and FAISS as vector databases
• 💬 Ollama LLM with the Mistral model for generating answers

---

📁 Document Files

Your documents go inside a folder named documents/:

---

⚙️ Setup Instructions

✅ Prerequisites

• Python 3.8+
• Ollama with mistral model (ollama run mistral)
• Rust (needed for Chroma on Windows): https://rustup.rs/
• Git (for cloning)

📦 Install Dependencies

git clone https://github.com/your-username/langchain-qa-chroma-faiss.git
cd langchain-qa-chroma-faiss
pip install -r requirements.txt

python main.py

1: Continue 
2: Exit 
Enter Your Question: What is payment?

Answer: Payment refers to the transfer of money or value from one party to another, typically in exchange for goods, services, or to fulfill obligations.

📄 License This project is licensed under the MIT License.