This talk will showcase an innovative AI-enabled learning tool for a real-world use case. Hear how the UBC Cloud Innovation Centre (CIC) developed a cloud-based solution to systematically identify and address student knowledge gaps in real time using instructor-vetted course materials. The solution is designed to accommodate diverse learning preferences and helps to ensure an inclusive and enriching educational experience for all students. It leverages large language models (LLMs) and AWS technologies to analyze curriculum content and monitor students' learning journeys.

By integrating a pedagogically determined concept learning journey with LLM-driven insights, the platform delivers synchronous, targeted feedback on each concept. It effectively unites instructors' expertise with AI's round-the-clock availability. Students benefit from having, in essence, a learning assistant that is accessible at any time, capable of explaining concepts in multiple ways with varied examples, and able to direct them to concepts they should revisit. The solution also benefits instructors as it provides them with analytics that delve into the level of student engagement with the tool.

We will begin this session with a concise introduction to the key Generative AI concepts at the heart of this solution, particularly large language models and Retrieval-Augmented Generation (RAG). Then, we will delve into the specifics of how the cloud-based solution was developed. Finally, the session will include a testimonial from an instructor who has piloted the solution in their class and will share firsthand insights into its implementation and benefits. Through this talk, we aim to demonstrate how these cutting-edge technologies can transcend simple conversational applications to fundamentally transform curriculum evaluation, foster individualized learning paths, and ultimately benefit both educators and students.

Summit Speaker

Harshinee Sriram

Graduate Student Researcher, Applied Scientist Intern, The UBC Cloud Innovation Centre (CIC)

Harshinee Sriram is a Ph.D. candidate in Computer Science at the University of British Columbia, specializing in multimodal learning, self-supervised learning, signal processing, and explainable AI, with a focus on neurodegenerative diseases. Her research aims to advance the early diagnosis and progression modeling of conditions such as Alzheimer’s and Parkinson’s disease using deep learning and multimodal data sources.

She has been recognized with multiple prestigious awards, including the UBC Advanced Machine Learning Training Network (AML-TN) Funded Fellowship, the 2024 BPOC Graduate Excellence Award, and the President’s Academic Excellence Initiative PhD Award (2021-2024). Her work has been published in leading conferences such as ML4H, IJCAI, ICMI, and UMAP, where she has also presented research on Alzheimer’s disease classification using deep learning on eye-tracking data and the role of personalized AI explanations in intelligent tutoring systems.

Harshinee has extensive experience in applied AI research through her role at the UBC-AWS Cloud Innovation Centre, where she has developed classical and generative AI solutions, retrieval-augmented generation pipelines, and machine learning-based analytics tools to address community-facing challenges.

Summit Speaker

Dr. Michael Jerowski

Education Technologist and Sessional Lecturer, University of British Columbia

Dr. Micheal Jerowsky is a community-based researcher specializing in the application of emerging technologies in environmental education. He is committed to integrating innovative tools and platforms into the classroom to enhance students’ digital literacy and prepare them for success in today’s world. Dr. Jerowsky believes that when applied effectively, these technologies can foster more flexible and accessible learning environments, enabling students to learn at their own pace and in ways that best suit their individual needs. Currently, he serves as an Educational Technologist with UBC Arts ISIT and a Sessional Instructor in UBC’s Geography Department.

Technology Track