BookWise - AI Chatbot for “Getting Started | Chai aur Docs”

A lightweight, retrieval‑augmented Streamlit chatbot that indexes content, stores embeddings in Qdrant, and answers questions grounded in the source with links back to the exact page/section.

This project is configured as a chatbot for the website “Getting Started | Chai aur Docs”

Tech Stack

• UI: Streamlit.
• Retrieval: LangChain (loaders, splitters) + Qdrant vector store.
• Models: OpenAI chat + embeddings (configurable in code).

Quick Start

Prerequisites: Python 3.9+, Git; Docker if running Qdrant locally.

1. Clone and install:

• git clone https://github.com/AnkitAnandMastery17/BookWise
• cd BookWise
• python -m venv .venv

2. Start your virtual environment

3. Install all the required files

   pip install -r requirements.txt

4. Configuration

• Create a folder .streamlit
• Inside the folder create a file secrets.toml.
• Add OPENAI_API_KEY = "<Your Api Key>"
QDRANT_URL = "http://localhost:6333"

5. Run Qdrant locally via Docker:

   docker compose up -d

6. Run python indexing.py

   To confirm your qdrant is running. visit http://localhost:6333/dashboard. Go to collections, you will see a new collection named BookWise.

8. Run the app streamlit run main.py
9. Go to http://localhost:8501