iSage is an AI-powered (RAG) legal companion designed to provide international students with trusted and up-to-date immigration guidance directly from official U.S. government sources.
Key Features:
- Accurate and Reliable: iSage leverages LightRAG, a simple and fast Retrieval-Augmented Generation system, to enhance large language models (LLMs) with knowledge from official U.S. government sources. This ensures that the information provided is accurate, reliable, and up-to-date.
- User-Friendly: iSage offers a user-friendly interface (accessible via Streamlit) that allows international students to easily access information and get their questions answered.
- Personalized Guidance: iSage can provide personalized guidance based on individual student circumstances, such as visa status, immigration history, and academic plans.
Technology:
- LightRAG: A lightweight and efficient RAG system that integrates external knowledge sources with LLMs.
- Learn more about LightRAG: https://github.com/HKUDS/LightRAG
- Large Language Models (LLMs): Powered by advanced LLMs to generate human-like text, answer questions, and translate languages.
Vision: iSage aims to empower international students with the knowledge and confidence they need to navigate the complexities of the U.S. immigration system. By providing easy access to reliable information and personalized guidance, iSage strives to reduce confusion and anxiety for students and help them achieve their academic and professional goals.
Disclaimer: iSage is an informational tool and should not be considered legal advice. Please consult with a qualified immigration attorney for any specific legal questions or concerns.
Get Started:
- Visit the iSage application: https://immigrant-student-advisory-engine.streamlit.app/
- Explore the features: Interact with the AI assistant, access helpful resources, and get answers to your questions.
Contribute: We welcome contributions to the iSage project. If you are interested in contributing code, improving documentation, or providing feedback, please contact us.
make create-venv
make setup
make run-streamlit