DataMate 🤖

DataMate is an interactive chatbot application designed to help users query and retrieve information from CSV files. Leveraging Google Generative AI, FAISS, and LangChain, DataMate ensures accurate, context-aware responses to your data-related questions.

Features

• CSV File Support: Seamlessly upload and process multiple CSV files.
• Conversational Interface: Query your data through an intuitive chat interface.
• Data Embeddings: Efficiently stores and retrieves data using FAISS.
• Google Generative AI: Provides precise, natural language responses.
• User-Friendly UI: Built with Streamlit for a clean and responsive interface.

---

Prerequisites

• Python 3.8 or higher
• Google API key for Generative AI
• Internet connection

---

Installation

1. Clone the repository:

   git clone https://github.com/Nomaanrizvi/DataMate-a-csv-chatbot.git
   cd DataMate-a-csv-chatbot

2. Install dependencies:

   pip install -r req.txt

3. Configure your environment:

   • Create a .env file in the project root directory with your Google API key:

   GOOGLE_API_KEY=your-google-api-key

---

Usage

1. Run the application:

   streamlit run main.py

2. Open the application in your browser (default: http://localhost:8501).

3. Upload CSV files using the sidebar and click "Submit & Process."

4. Start chatting and querying your data interactively.

---

How It Works

1. Data Parsing: Extracts structured data from uploaded CSV files.
2. Text Chunking: Breaks large datasets into manageable text chunks.
3. Embedding Storage: Generates embeddings using Google Generative AI and stores them in FAISS for efficient searches.
4. Question Answering: Uses a conversational chain model to provide relevant responses based on your questions.

---

File Structure

• main.py: Main application script for DataMate.
• req.txt: Contains the list of required Python libraries.
• .env: Environment configuration file (user-provided).
• faiss_index/: Stores FAISS vector indexes (created during runtime).

---

Future Enhancements

• Support for additional data formats like Excel and JSON.
• Improved conversational AI for more complex queries.
• Integration with visualization tools for better data insights.

---

Contributing

We welcome contributions! Feel free to open issues or submit pull requests to improve DataMate.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.