Project Details

[Return to Previous Page]

Penn State Advising LLM Integrated Proposal

Company: PSU Computer Science and Engineering

Major(s):
Primary: CMPSC

Non-Disclosure Agreement: NO

Intellectual Property: NO

This initiative aligns with Penn State’s strategic priorities in student success, digital transformation, and responsible AI integration. This proposal builds upon insights gathered from Computer Science and Engineering academic advisers across Penn State, whose feedback continues to guide the design and priorities of this initiative. Abstract: Penn State’s advising system is essential to student success, yet advisers especially in Computer Science and Engineering are overwhelmed by large caseloads, high email volume, and complex workflows in systems like LionPATH and Starfish, resulting in delayed appointments, inconsistent policy interpretation, and students turning to peers for guidance. The proposed Penn State Advising LLM Framework offers a practical, future-focused solution by providing a policy-driven, AI-assisted environment that supports both advisers and students through consistent, validated guidance, automated retrieval of official policies, and 24/7 access to academic support. By integrating securely with Penn State’s existing systems, incorporating human-in-the-loop oversight, and applying FERPA-aligned governance, this framework aims to reduce adviser workload, improve graduation timelines, empower underserved and first-generation students, and enhance retention across the university ultimately strengthening Penn State’s advising ecosystem while preserving trust, equity, and academic rigor. Main Objective To design, pilot, and evaluate the Penn State Advising LLM Framework, a policy-driven system that enhances advising consistency, automates routine tasks, and strengthens support for students and advisers through secure, integrated, and data-informed solutions. Specific Aims • Design a policy-driven advising framework aligned with Penn State’s approved academic policies across colleges and departments. • Pilot the system to automate routine advising tasks such as petition formatting, degree audit guidance, and common student inquiries to reduce adviser workload. • Evaluate system performance, usability, and accuracy through adviser and student feedback in real advising contexts. • Refine the model through collaboration and governance, maintaining transparency and policy compliance across all units.