🤖 Basic RAG Chatbot

A simple Retrieval-Augmented Generation (RAG) Chatbot built using LangChain, Ollama, ChromaDB, and Streamlit.

This chatbot answers user queries based on document context using vector search + LLM reasoning.

---

🚀 Features

• 🔍 Semantic search using vector embeddings
• 🧠 Context-aware answers using RAG pipeline
• 💬 Conversational interface with chat history
• ⚡ Local LLM using Ollama (no API cost)
• 📊 Fast retrieval using Chroma vector database
• 🎯 Accurate answers grounded in provided documents

---

🏗️ Tech Stack

• Backend Framework: LangChain
• LLM: Ollama (LLaMA 3.2)
• Embeddings: Ollama Embeddings (nomic-embed-text)
• Vector Database: ChromaDB
• Frontend: Streamlit
• Language: Python

---

⚙️ How It Works (RAG Flow)

1. User enters a question
2. Query is converted into embeddings
3. ChromaDB retrieves relevant document chunks
4. Context is combined with the question
5. LLM (Ollama) generates final answer

---

🛠️ Installation

1. Clone the repository

git clone https://github.com/Sunanda01/Basic_RAG_Chatbot.git
cd Basic_RAG_Chatbot

2. Create Virtual Environment

python -m venv venv
source venv/bin/activate     # Mac/Linux
venv\Scripts\activate        # Windows

3. Install dependencies

pip install -r requirements.txt

4. Install & Run Ollama

Run required models:

ollama pull llama3.2
ollama pull nomic-embed-text

5. Run the application

# Document Injestion 
python injest.py

# Run UI
streamlit run home.py

---

💬 Usage

Open browser at http://localhost:8501
Ask questions based on your documents
Chatbot will respond using retrieved context

---

📸 UI

---

▶️ Demo

chatbot-demo.webm