Customer-Assistant-Chatbot

A Retrieval-Augmented Generation (RAG) based LLM chatbot designed to answer technical customer service questions using a structured internal knowledge base.

The chatbot retrieves relevant policies from a PostgreSQL database, performs vector similarity search using PgVector, and generates accurate responses using the Groq LLM API.

Instructions

1. Clone the Repository

git clone https://github.com/yourusername/Customer-Assistant-Chatbot.git
cd Customer-Assistant-Chatbot

---

2. Create a Python Virtual Environment

python -m venv venv

Activate the environment:

Linux / macOS

source venv/bin/activate

Windows

venv\Scripts\activate

---

3. Install Dependencies

pip install -r requirements.txt

---

4. Install PostgreSQL

Install PostgreSQL if it is not already installed.

Then create a new database:

CREATE DATABASE customer_chatbot;

---

5. Enable PgVector Extension

Connect to the database and enable the extension:

CREATE EXTENSION IF NOT EXISTS vector;

---

6. Create the Documents Table

CREATE TABLE documents (
    id INTEGER PRIMARY KEY,
    category VARCHAR(50) NOT NULL,
    prompt TEXT NOT NULL,
    response TEXT NOT NULL,
    embeddings VECTOR(1536)
);

7. Add Environment Variables

Create a .env file in the project root:

# Database
DB_HOST= your_db_host
DB_PORT= your_db_port
DB_NAME= your_db_name
DB_USER= your_db_user
DB_PASSWORD= your_db_password

# Groq
GROQ_API_KEY= your_groq_api_key

8. Generate and Store Embeddings

Run the embedding script to convert the 50 policies into vectors and store them in PostgreSQL.

python embeddings/embeddings.py

9. Start the Chatbot

Run the chatbot:

python streamlit "src\app.py"