17630-StudentGuideAi

Problem Statement

New and current Carnegie Mellon University (CMU) students face challenges navigating the complex academic ecosystem. An overwhelming volume of academic policies, like understanding cross-registration, involves numerous rules and forms​. Information is spread across multiple data sources in university policy PDFs, departmental FAQs, and college handbooks. The generic Q&A chatbots fall short for CMU-specific queries, and the static FAQ pages lack interactive guidance, while general-purpose large language models aren’t tailored to CMU’s latest policies and may hallucinate university-specific rules. For example, a generic chatbot might confidently give the wrong course add deadline or an outdated graduation requirement. No centralized tool guided students from orientation to graduation on CMU-specific academic matters. These gaps motivated the creation of “StudentGuideAi,” an AI-powered assistant that provides reliable answers and streamlines the student experience.

Create .env with the following variable:

1. OPENAI_API_KEY=
2. OPENAI_ORG=
3. OPENAI_PROJECT_ID=

Experimentation and RAG Workflow?

• We have python notebooks in the folder.
• [WIP] We're constantly improving our experimentations.

How to Run the Project?

• Automated Setup:

1. Run the script using ./setup_and_run.sh

• Manual Setup:

1. Create a local virtual environment using python3 -m venv prompt-project
2. Activate the environment using source prompt-project/bin/activate
3. Install the necessary dependencies using pip install -q -r requirements.txt
4. Build the RAG using rag.py
5. Run the project on localhost using python main.py

Documentation

• We have drafted a slideshow (Click here) and a report (Click here) about the project.