# AI Driven Customer Support System

This project implements an intelligent customer support system leveraging Streamlit and a fine-tuned Llama large language model (LLM) from Hugging Face. It empowers users to interact with the system through text input, speech recognition, and tailored responses based on the chosen assistance type (Agriculture, Healthcare, Education).

Key Features:

• Seamless User Interaction:
  • Text input for clear queries.
  • Speech recognition for hands-free convenience.
  • Response word limit selection for concise or detailed answers.
  • Radio button selection for focused assistance type.
• Enhanced Response Generation:
  • Pre-trained Llama model fine-tuned on a relevant dataset (https://huggingface.co/datasets/PRAKALP-PANDE/PSP-agricultureQnA-1k-unique).
  • Caching mechanism for efficient retrieval of existing responses.
• Accessibility:
  • Text-to-speech functionality for reading responses aloud.

Getting Started:

1. Prerequisites:

   • Python 3.x (https://www.python.org/downloads/)
   • Git version control system (https://git-scm.com/)
   • Anaconda package manager (https://www.anaconda.com/products/distribution)

2. Clone the Repository:

   git clone https://github.com/your-username/AI-Driven-Customer-Support-System.git

   (Replace your-username with your actual GitHub username)

3. Set Up a Conda Environment (Optional but Recommended):

   • Create a virtual environment to isolate project dependencies and avoid conflicts with other projects:

     cd AI-Driven-Customer-Support-System
     conda create -n dcss_env python=3.8  # Adjust Python version as needed
     conda activate dcss_env

4. Install Required Libraries:

   • The requirements.txt file specifies all necessary dependencies:

     pip install -r requirements.txt

Project Structure:

AI-Driven-Customer-Support-System/
├── app.py                     # Main application script
├── models/                     # Folder to store the downloaded LLM model
│   └── PSP-AI-DCSS-finetune.bin  # (Download from https://huggingface.co/PRAKALP-PANDE/PSP-AI-DCSS-finetune)
├── ReadMe.md                   # This file (project documentation)
├── requirements.txt            # List of required Python libraries
└── .gitignore                   # Specifies files to exclude from version control

Model Download and Usage:

1. Download the Pre-trained Model:
   • Access the LLM model from the provided link (https://huggingface.co/PRAKALP-PANDE/PSP-AI-DCSS-finetune).
   • Download the model file (e.g., PSP-AI-DCSS-finetune.bin).
2. Move the Model File:
   • Place the downloaded model file inside the models folder within the project directory.

Running the Application:

1. Activate the Conda Environment (if you created one):

   conda activate dcss_env

2. Start the Streamlit Development Server:

   streamlit run app.py

3. Access the Application:

   • Open http://localhost:8501 in your web browser.

Usage:

1. Interact with the System:
   • Type your query in the text input field or use speech recognition.
   • Select the desired word limit for the response.
   • Choose the type of customer assistance you need (Agriculture, Healthcare, Education).
   • Click "Generate" to receive a response from the model.
2. Response Generation:
   • The system will leverage the fine-tuned LLM to generate a tailored response based on your input and chosen assistance type.
   • Caching is implemented to retrieve existing responses efficiently.
3. Text-to-Speech (Optional):
   • Click "Listen to the Response" to have the system read the response aloud using text-to-speech.

Built With:

• Streamlit (web framework)
• langchain (LLM interaction)
• PyMongo (optional, for database integration)
• SpeechRecognition (speech-to-text)
• pyttsx3 (text-to-