The Query asssit is a robust solution designed to assist students with administrative queries using RAG and LLMs. It serves as a virtual advisor, providing timely and accurate information, making the administrative process more efficient and student-friendly.
- Uses Mistral large model/Groq: Groq helps in faster inference speeds, whilst Mistral gives a slightly accurate answer, based on user's need.
- Information Retrieval: Utilizes RAG to fetch relevant information from a vast knowledge base.
- User-Friendly Interface: Easy to interact with, providing a seamless user experience.
- Multi-Lingual Support: Responds to the user query in their language
- Ticket Generation: Summarizes the conversation and generates a support ticket to the admin team automatically if the user needs more help
- Python 3.7 or higher
- pip (Python package installer)
- Clone the repository:
git clone https://github.com/Roshan-Velpula/Mistral-Hackathon-2024.git
- Create and activate a virtual environment
In MacOS:
virtualenv .venv source .venv/bin/activate - In Windows:
virtualenv .venv .venv\Scripts\activate
- Install the required packages(Note:change directory to 'Project folder')
pip install -r requirements.txt
- Run the Streamlit application
streamlit run CampusBOLT.py