What is it about?
On many campuses, a single academic advisor is responsible for hundreds of students, so it's hard for students to get timely, accurate help planning their courses. This work introduces Aurora, an AI advising assistant built to ease that strain. Aurora pairs a language model that explains advice in plain, conversational words with a separate logic system that checks every recommendation against the school's real requirements (things like course prerequisites and credit limits). That second step is the key: rather than just producing answers that sound right, Aurora verifies each suggestion against official program rules before a student ever sees it, and it's designed to say so when it doesn't have enough information to answer. The result is course guidance that students and advisors can actually trust.
Featured Image
Photo by Vitaly Gariev on Unsplash
Why is it important?
AI chatbots are now everywhere, but they share a well-known flaw: they can state wrong information confidently and even invent things that don't exist (often called hallucination). In academic advising, that is a real risk: a single bad recommendation can suggest a course a student isn't eligible for, push them over a credit limit, or delay their graduation. Aurora is built to prevent this. Every recommendation is automatically checked against the institution's actual rules, and each answer can be traced back to the exact requirement behind it. In testing, Aurora's recommendations matched expert advisors' answers far more closely than a standard AI chatbot's did. Just as important, it's meant to support human advisors rather than replace them (by handling routine, rule-heavy questions so advisors have more time for the personal guidance only a person can give).
Perspectives
Aurora grew out of a frustration many students know firsthand: when you're trying to plan your path to graduation, it can be hard to get clear, timely answers. What excites me most about this project is that we didn't just build something that responds quickly: we built something that gives answers you can check! To me, that's the heart of trustworthy AI: not asking people to take the system's word for it, but showing the reasoning behind every recommendation. I'm proud of what our team accomplished, and I hope it points toward AI that makes academic support more dependable and accessible for students, while keeping human advisors at the center.
Lorena Amanda Quincoso Lugones
Florida International University
Read the Original
This page is a summary of: Aurora: Neuro-Symbolic AI Driven Advising Agent, March 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3748522.3779850.
You can read the full text:
Contributors
The following have contributed to this page
