🛍️ Retail RAG-based QA Bot

A Retrieval-Augmented Generation (RAG) powered chatbot tailored for the retail industry. This bot allows users to ask questions related to return policies, shipping, refunds, and more. It dynamically retrieves relevant information and generates answers using powerful language models.

---

🚀 Key Features

• 🔍 Contextual Retrieval: Vector search from retail policies.
• 🤖 Dual Model Support:
  • Gemini Pro (Google) — via API for powerful cloud-based generation.
  • FLAN-T5 (HuggingFace) — lightweight offline model for local use.
• 🔄 Model Switcher in sidebar for dynamic backend selection.
• 📚 Streamlit Interface for clean and interactive UI.
• ⚡ Fast response, supports multiple queries.

---

🧠 Tech Stack

• Frontend/UI: Streamlit
• Retrieval Engine: SentenceTransformers, FAISS
• Generation Backends:
  • google.generativeai (Gemini)
  • transformers + Flan-T5

---

🖼️ Demo Screenshots

🟢 Gemini Pro Output

"Yes, you can return a furniture item if it was damaged in transit."

🔵 HuggingFace FLAN-T5 Output

"not eligible for return unless damaged in transit"

---

📂 Project Structure

retail-rag-qa-bot/
├── app.py
├── requirements.txt
├── README.md
├── assets/
│   ├── gemini_output.png
│   └── hf_output.png

---

🔧 Setup Instructions

🔹 1. Clone the Repo

git clone https://github.com/yourusername/retail-rag-qa-bot.git
cd retail-rag-qa-bot

🔹 2. Install Dependencies

pip install -r requirements.txt

🔹 3. Add API Key for Gemini

• Create a .env file or set environment variable:

export GOOGLE_API_KEY="your-api-key"

🔹 4. Run the App

streamlit run app.py

---

🧪 Sample Questions

• Can I return a damaged furniture item?
• What is the refund timeline?
• Where do you ship in India?
• Is COD available for ₹10,000?

---

🙌 Acknowledgements

• Google Gemini Pro (generativeai)
• HuggingFace Transformers
• Streamlit Community

---

📬 Contact

Made with ❤️ by Divyansh Gautam

LinkedIn | GitHub