ChatBot-Snowflake

Overview

ChatBot-Snowflake is an advanced AI-powered chatbot designed to seamlessly interact with Snowflake databases, enabling intelligent data exploration, analysis, and visualization through natural language queries. This project leverages cutting-edge AI models and cloud computing capabilities to provide real-time insights from structured datasets.

The system comprises two Streamlit applications:

1. AI-Driven Chatbot for Dataset Interaction – An intelligent assistant that allows users to query structured data using natural language and receive AI-generated insights.
2. Cortex Analyst for Interactive Data Visualization – A powerful visualization tool that integrates with Snowflake Cortex to provide data-driven insights through interactive plots and analytics.

Key Features

Application 1: AI-Powered Chatbot for Dataset Interaction

• Conversational AI Interface – Enables users to query datasets using natural language.
• Snowflake Snowpark Integration – Retrieves and preprocesses large datasets with high efficiency.
• LangChain-Powered Processing – Uses RecursiveCharacterTextSplitter for structured data handling.
• AI-Driven Response Generation – Utilizes Snowflake Cortex LLMs for intelligent data-driven responses.
• Multi-Model AI Support – Allows switching between advanced models such as llama3.2-3b, claude-3-5-sonnet, mixtral-8x7b, and snowflake-arctic.

Application 2: Cortex Analyst for Data Visualization

• Dynamic Data Retrieval – Connects to Snowflake databases to fetch and analyze structured data.
• AI-Powered Data Interpretation – Leverages Cortex Analyst to generate insights from complex datasets.
• Interactive Visualizations – Uses Plotly for generating intuitive and interactive data charts.
• Conversation Memory & Context Awareness – Retains session history for a more intuitive user experience.
• Custom Semantic Model Selection – Allows users to choose from predefined model paths for optimized performance.

Installation & Setup

Prerequisites

Ensure the following dependencies are installed before running the chatbot:

• Python 3.8+
• Streamlit
• Snowflake Connector for Python
• LangChain
• Pandas
• Plotly
• A secrets.toml file for Snowflake credentials

Installation Steps

1. Clone the repository:

  git clone <repo-url>
  cd <repo-directory>

2. Create a virtual environment and activate it:

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

3. Install dependencies:

   pip install -r requirements.txt

4. Set up secrets.toml for Snowflake credentials:

toml
   [snowflake]
   account = "your_snowflake_account"
   user = "your_username"
   password = "your_password"
   warehouse = "your_warehouse"
   database = "your_database"
   schema = "your_schema"

5. Run the Streamlit app:

   streamlit run streamlit_app.py  # For AI-powered dataset interaction
   streamlit run cortex_analyst.py  # For Cortex Analyst and visualization