mamd555 / llm-powered-bot

Utilizing the Advanced Capabilities of GPT and Large Language Models (LLMs), this bot is expertly crafted to facilitate interactive engagement with enterprise data. Designed for seamless integration and efficient data handling, it represents a leap forward in enterprise-level data interaction and management.

LLM-Based Power Virtual Agent and Power Automate Chatbot

This project leverages the power of OpenAI's GPT models using Langchain to provide a robust chatbot for Power Virtual Agent and Power Automate platforms. Our solution enriches backend operations with advanced NLP capabilities to process user input effectively and generate meaningful responses.

Features

• LLM Backend: Uses Azure deployment of OpenAI's GPT models.
• Data Preprocessing: Advanced data preprocessing module to optimize user input for GPT model consumption.
• Prompt Engineering: Utilizes prompt templates to guide the GPT models and produce tailored responses suitable for chatbot interactions.
• Integration with Power Platforms: Designed to be seamlessly integrated with Microsoft's Power Virtual Agent and Power Automate platforms.

Prerequisites

1. Python 3.11 or higher.
2. OpenAI Azure Deployment.
3. Microsoft's Power Virtual Agent and Power Automate platforms account.

Setup

1. Clone the Repository:

   git clone [repo_link]
   cd [repo_directory]

2. Install Dependencies:

   pip install -r requirements.txt

3. Environment Variables: Copy the .env.example to .env and update the variables including the OpenAI API Key, API base URL, etc.

4. Run the Backend:

   python api.py

Usage

1. Start the Streamlit UI:

   python app.py

2. Navigate to the provided URL in your browser. Input the desired genre, characters, and news text. Click on "Begin!" to get the novel based on your inputs.

Contributing

We welcome contributions! Please create an issue to discuss the proposed changes or submit a pull request.

License

MIT License