Notion-Chroma-Chatbot

A Streamlit chatbot application that integrates Notion, ChromaDB, and LangChain. The chatbot leverages pre-trained embedding models to enable users to query information efficiently from Notion databases.

Features

• Document Parsing: Load and process documents directly from Notion using NotionDBLoader.
• Chunking: Split large documents into manageable chunks for efficient retrieval.
• Chroma Vector Store: Store embeddings and metadata in ChromaDB for similarity search.
• Streamlit Frontend: An interactive web-based interface for chatting with the chatbot.
• Open Source Models: Uses the lightweight all-MiniLM-L6-v2 embedding model for fast and accurate results.

---

Installation

1. Clone the Repository

git clone https://github.com/jabahm/notion-chroma-chatbot.git
cd notion-chroma-chatbot

2. Create a Virtual Environment

python3 -m venv venv
source venv/bin/activate  # For Windows: venv\Scripts\activate

3. Install Dependencies

pip install -r requirements.txt

---

Usage

1. Configure Environment Variables

Create a .env file in the project root with the following variables:

NOTION_INTEGRATION_TOKEN=<Your Notion Integration Token>
NOTION_DATABASE_ID=<Your Notion Database ID>

2. Load Data from Notion to Chroma

Run the notion_to_chroma.ipynb notebook or convert it to a Python script. This will:

• Load documents from your Notion database.
• Chunk the documents into smaller pieces. -Generate embeddings and save them to ChromaDB.

3. Start the Streamlit Chatbot

streamlit run app.py

---

Using Docker

You can containerize the application using the provided Dockerfile.

1. Build the Docker Image

docker build -t notion-chroma-chatbot

2. Run the Docker Container

docker run -p 8501:8501 --env-file .env notion-chroma-chatbot

The chatbot will be accessible at http://localhost:8501.