Company Chatbot

A Retrieval-Augmented Generation (RAG) based chatbot that answers company-related queries using information stored in company.txt. The chatbot is optimized to run with a single local language model, providing a simpler and more reliable experience.

Features

• Company-specific question answering
• Uses a single local language model
• Optimized for faster loading and lower resource usage
• Supports local model configuration through .env
• Improved model compatibility for tool calling
• Validated with multiple test cases for consistent responses

Clone the Repository

git clone https://github.com/AgentRoomAI/chat-agent.git
cd chat-agent

2. Create a Virtual Environment

python -m venv venv

3. Activate the Virtual Environment

Windows

venv\Scripts\activate

Linux/macOS

source venv/bin/activate

4. Install Dependencies

pip install -r requirements.txt

5. Download the Model

The application is configured to use a single default local model for improved performance and simplified model management.

python download_models.py

If you want to use a different locally downloaded model, set LOCAL_MODEL_PATH in a .env file to the model directory path before running the application.

Run the Chatbot

Start the chatbot using:

python chatbot.py

Usage

1. Add company information to company.txt.
2. Run the chatbot.
3. Ask questions related to the company.
4. The chatbot answers using the information available in the company document and the configured local model.

Recent Updates

• Refactored the chatbot to use a single local language model instead of loading multiple models.
• Integrated a model that is better suited for tool calling.
• Simplified the model loading process for improved efficiency.
• Validated the chatbot with multiple test cases to ensure stable and reliable responses.

Note

If information is not available in company.txt, the chatbot will return:

Sorry, I don't have that information.

Author

Aman V