AI-driven adjustments ensure that if students change majors, drop courses, or face obstacles, their plan recalibrates in real time.

Benefits include enhanced registration satisfaction, better predictive course demand data for departments, and reduced time advisors spend on transactional planning

This will allow advisors to focus on listening and coaching rather than administrative note-taking.

Potential “copilot” features could recommend institutional resources or highlight risk signals (e.g., repeated course withdrawals, unearned credits).

24/7 student support with policy-vetted, contextual answers (registration, holds, deadlines).

Could operate in a Human-in-the-Loop (HITL) model to escalate to human advisors for complex or ambiguous cases.

Creates a scalable hybrid system balancing automation with human judgment.

Perceived Benefits: reduces routine queries, enhances access, aligns with enterprise IT principles (secure, accurate, scalable).

This approach builds on DataKind’s demonstrated success at John Jay College, where predictive modeling contributed to a 32-point increase in the senior graduation rate.

Outputs would take the form of “support scores” that are used internally by advisors to prioritize outreach without stigmatizing students.

A USC-specific model could reveal risk factors tied to program pathways, enabling proactive course recommendations, referrals, or interventions.

Designed by Travis Gardner using ChatGPT