This talk presents the current status and evolving vision of the Augmented University Real-time Assistant (AURA), our multi-semester effort to integrate AI into teaching and learning in universities. This talk will cover the current platform, including role-based dashboards for teachers and students, and an assessment pipeline that allows instructors to upload course materials and rubrics so that AURA can generate multi-modal assessments and automatically grade them with consistent, rubric-based feedback.

This talk will also present the asynchronous side of AURA, where students can launch personal learning sessions and interact with an AI instructor. Work in progress includes connecting a classroom microphone to diarize all human-teacher speech into the course material database for personal AI tutor chatbots. Lastly, the roadmap ahead will be discussed: richer pedagogical scaffolding for students, adaptive assessments that adjust difficulty based on performance for enhanced personalization, and stronger guardrails against cyber threats such as malicious prompt injection, as well as a discussion of current limitations.

Dr. Samer Saab Jr. is an Assistant Professor of Computer Engineering at the Lebanese American University (LAU). He received both his B.S. and Ph.D. degrees in Electrical Engineering from the Pennsylvania State University. Before joining LAU, he spent three years in California as a Senior Staff Machine Learning Scientist at Enevate Corporation.

His research interests include optimization in deep neural networks, task decomposition for multi-LLM agent systems, multi-LLM consensus dynamics, and context compression for large language models.