Poster Gallery Walk
Bringing together the people and ideas shaping what's next for higher education.
October 14-16, 2026
Luskin Conference Center, UCLA
An invitation-only convening
Bringing together the people and ideas
shaping what's next for higher education.
October 14-16, 2026
Luskin Conference Center, UCLA
An invitation-only convening
PROGRAM
Conversations and ideas reshaping higher education
Explore five mainstage panels, sixteen breakout sessions across five Learning Tracks, a Poster Gallery Walk, and more. The program brings together ideas, expertise, and innovative work from Learning Lab’s grantee network and leaders across California’s three public higher education systems.
Poster Gallery Walk
Thurs, OCT. 15 | 2:15 – 3:00 pm
Explore innovative projects and emerging ideas from across the Learning Lab community and beyond. Presenters will be available throughout the Poster Gallery Walk to discuss their work and answer questions.
Building Connected Data Science Pathways with the PIPE-LINE Framework
Programs for Institutional Pathway Engagement–acceLerating INfrastructure and Education (DATA SCIENCE)
Jessica Jaynes, CSU Fullerton
Data science education is growing rapidly, but student success depends on strong connections across disciplines, institutions, and communities. This poster introduces the PIPE-LINE Framework, a collaborative approach to building coordinated data science pathways. Drawing on work in cross-disciplinary curriculum development, articulation, faculty communities of practice, undergraduate research, and statewide partnerships, we share practical strategies for connecting students’ educational experiences across campuses and communities. Visit our poster to explore how the framework can strengthen data science education and broaden participation across California and beyond.
One Curriculum, Multiple Languages: AI-Assisted Translation and Verification for Data Science
Collaborative Design of Interdisciplinary, Inclusive, Introductory Data Science Course Modules (DATA SCIENCE)
Indar Freitas, UC Merced
This poster highlights our process for adapting UC Berkeley’s Data 6 curriculum to different programming environments, helping to lower barriers to course adoption and supporting the growth of data science programs across diverse educational settings. Using an AI-assisted approach, we converted UC Berkeley’s existing, custom Python-based datascience.Table library to both Polars (alternative Python dataframe library) and the tidyverse (open-source R packages). We developed translation guides that map commands across frameworks, then used AI to convert assignments while preserving instructional context. To support quality, converted materials were reviewed by AI “student” and “teacher” verifiers, checked by human translators, and tested to ensure all code functioned as intended. The resulting open-source curriculum will provide institutions with flexible, ready-to-deploy options for teaching data science in different contexts.
AI, Memory, and Civic Learning: Listening Across Generations with Hibiki-AI
The Japanese American Incarceration Testimony Interactive Q&A Project (AI FAST)
Rowan Elansary, Fresno State
Hanayo Oya, Fresno State
Thanos Panagopoulos, Fresno State
Jermaine Steele, Fresno State
How can artificial intelligence help students meaningfully engage with firsthand accounts of the Japanese American incarceration during World War II? Hibiki-AI explores an interactive approach to preserving historical memory, encouraging close listening to survivor voices. This poster and demonstration introduces an AI-supported tool that helps students explore recorded testimonies, identify historical themes, formulate thoughtful questions, and connect individual experiences with broader issues of civil liberties, racism, and resistance. Rather than replacing survivors’ voices, Hibiki-AI guides students to interact with and ask questions to an AI-supported virtual survivor persona, leading students back to original testimony and verified historical sources. You’re invited to experience the tool, examine sample classroom activities, and consider the possibilities and ethical challenges of using AI with sensitive historical materials. Hibiki-AI offers a model for combining technology, media literacy, and responsible historical education while keeping human experiences and authentic testimony at the center.
Centering Student Experience and Identity in Data Science Program Evaluation
Building Inclusive and Collaborative Foundations in Data Science (Data Science)
John Louis-Strakes Lopez, UC Berkeley
Jonathan Pedroza, UC Berkeley
Kseniya Usovich, UC Berkeley
Data science is inherently interdisciplinary, so why evaluate data science programs with a single method? This poster presents three complementary approaches used to evaluate a community college’s implementation of an introductory data science course adapted from UC Berkeley. We examine traditional “objective” outcomes such as grades, then layer in “subjective” student experience survey data that voices how learners actually experienced the course. Then, drawing on educational psychology, we explore how beliefs about math and programming abilities shape emerging identities and self-concept in data science. Together, these lenses provide a richer understanding of teaching and learning than any one measure alone, generating context-specific insights for community colleges. If you’re looking for new possibilities for evaluating STEM education beyond grades, come talk to us.
Cal-ICOR: Statewide Infrastructure for Equitable Data Science Education
Partnership for Establishing a California Interactive Computing Open Resource (Cal-ICOR) (PARTNERSHIP)
John DeNero, UC Berkeley
Shane Knapp, UC Berkeley
Eric Van Dusen, UC Berkeley
Edwin Vargas Navarro, UC Berkeley
What does it take to bring modern computing into every California classroom, regardless of campus size or technical capacity? This poster introduces Cal-ICOR (California Interactive Computing Open Resource), a UC Berkeley-led initiative that expands access to data science education across the UC, CSU, and California Community College systems. Cal-ICOR addresses common barriers through four pillars: browser-based computing environments requiring no local installation, open and adaptable curricula, hands-on faculty support, and a statewide community of practice. Designed as durable educational infrastructure, Cal-ICOR currently connects 20 partner institutions and continues to grow. Come by to see how a shared JupyterHub platform is helping campuses move past technical barriers so instructors can focus on teaching, and students can focus on learning.
From Turntables to Transfer-Level Success: Culturally Relevant, Interdisciplinary Teaching in Umoja and Puente
*Not a Learning Lab funded project
Roberto Rubalcaba, San Diego City College
In Fall 2025, San Diego City College’s Umoja and Puente learning communities achieved 100% success in transfer-level mathematics with Honors. This presentation explores how culturally relevant pedagogy, high expectations, and high support helped students who had often struggled with math develop confidence, mastery, and leadership as peer teachers. As a case study, we highlight a collaborative Math and English lesson built around records, DJ culture, and hip-hop sampling. Students explore plagiarism and citation through album liner notes that credit sampled artists, while simultaneously investigating mathematical concepts embedded in music production, including percentages, periodic functions, function transformations, and weighted averages. The presentation includes a live demonstration of sampling and DJ techniques to illustrate these concepts in real time. By connecting academic content to students’ lived experiences and cultural knowledge, we demonstrate how interdisciplinary teaching promotes engagement, belonging, and exceptional student success.
SpartanAI: Scaling High-Quality Course Design with AI
SpartanAI: Elevating Course Design with Faculty and Student-Centric AI Agent (AI FAST)
Hriday Ampavatina, San José State
Barbra Burch, Quality Matters
Magdalini Eirinaki, San José State
Yingjie Liu, San José State
Great teaching deserves great support. SpartanAI is an AI-powered instructional design assistant that helps faculty create high-quality courses through evidence-based, actionable feedback. It demonstrates how purpose-built AI can expand instructional design support while preserving faculty expertise and pedagogical decision-making. Unlike general-purpose AI tools, SpartanAI applies instructional design principles and Quality Matters-informed practices to provide contextualized recommendations for course improvement. Visit our poster for a live demonstration, learn about our interdisciplinary design process, and discuss how responsible AI can strengthen faculty support, improve course quality, and scale instructional design capacity across higher education.
Curious to Willing to Ready: Fostering Change with Teamworthy Tasks
Expanding Equity and Access in Discrete Mathematics (Scaling Success)
Shandy Hauk, San Francisco State
Tim Hsu, San José State
Kathy Jensen, Evergreen Valley College
This poster explores how discrete mathematics instructors build capacity to move beyond lecture toward active, collaborative mathematical learning. Drawing on a 25-hour professional development workshop, we examine a progression from curiosity to willingness to readiness for instructional change. Participants engaged in hands-on learning, research-to-practice discussions, collaborative problem-solving, and lesson design, experiencing active learning as both students and teachers. One finding stood out: learning with colleagues matters. Participants consistently identified shared challenges and collegial conversations as the most valuable aspects of their six months of interaction. The ultimate lesson for us as designers is that readiness is not a fixed trait but something developed through faculty community, shared resources, experimentation, and adaptation. Visit our poster to discuss what helps instructional change take root and last.
Learning That Follows the Student: Curriculum-Wide AI Support for Engineering Education
Across-the-curriculum AI-based personalized assistants to support student success in engineering (AI GRAND)
Mingheng Li, Cal Poly Pomona
Sundararajan Venkatadriagaram, UC Riverside
GESTALT is a web-based system that allows integrated delivery of educational content, self-paced quizzes and context-specific AI chatbots to support mastery of domain knowledge among engineering students across the entirety of their major curriculum, including for transfer students. Recognizing that the engineering curriculum is sequential and tightly constrained through pre-requisites, student performance in prior classes is used to build a model of the their knowledge and to inform the level and nature of support that the student would need in subsequent classes. We use lessons from the science of learning, such as distributed practice, in conjunction with the student’s learning model to personalize the delivery for individual learners.
Enhancing Student Learning in Physics and Biology Through an AI-Powered Virtual Reality Classroom Assistant
Revolutionizing STEM Education: Integrating AI and Virtual Reality to Enhance Learning and Engagement (AI FAST)
Horace Crogman, CSU Dominguez Hills
Kwame Ehsun, Sierra College
This poster explores how combining artificial intelligence and virtual reality can make learning more interactive, personalized, and accessible. Faculty and researchers at California State University, Dominguez Hills and Sierra College have developed an AI-powered VR learning environment for introductory Physics and Biology courses. Students explore virtual labs and simulations while an AI tutor provides real-time explanations, feedback, and guidance tailored to individual learning needs. The project examines the impact of these tools on student understanding, engagement, confidence, and access to learning. Early results suggest that AI-enhanced VR can help students visualize complex concepts and learn more actively. Visit our poster to experience the technology and discuss its potential across disciplines.
Data Science Without Silos: Building Interdisciplinary Communities for Learning and Discovery
Building Data Science Communities for Improving Student Success (Data SCIENCE)
Marcella Gomez, UC Santa Cruz
Pedro Morales-Almazan, UC Santa Cruz
What becomes possible when educators and researchers connect across disciplines and institutions? This poster highlights how the Educational Group for Enrichment in Data Science fosters professional development, interdisciplinary curriculum innovation, and cross-institutional research collaboration. We share how these efforts break down disciplinary silos, support early-career researchers, and build sustainable communities for data science learning and discovery. Visit our poster to learn more and join the conversation about creating stronger connections across fields and institutions.
Transforming Marketing Education with Curriculum-Wide AI Integration
Customizable AI Integration in Higher Education: Marketing Example (AI FAST)
Yufan Lin, Cal Poly Pomona
Fan Wang, Cal Poly Pomona
Generative AI is transforming business education, requiring students to critically evaluate AI-generated information while preparing for AI-enabled workplaces. This poster presents a customizable, curriculum-wide framework for integrating AI into undergraduate marketing education through scaffolded learning experiences that span foundational courses to industry-sponsored capstone projects. Students develop AI literacy by comparing AI-generated and authentic data, applying AI to solve real business problems with community partners, and engaging with AI-powered learning supports that improve access for multilingual, first-generation, and working students. The framework combines authentic assessment, project-based learning, and inclusive instructional design with reusable AI-enhanced course modules, case studies, tutoring tools, and implementation resources.The courses built a student pipeline for an AI in Business Hackathon, where students solved real-world, sponsored challenges. Designed for adaptation across institutions with varying resources and AI expertise, this work demonstrates how AI can simultaneously strengthen workforce readiness, deepen learning, and reduce barriers to student success.
From Arabic to Business: AI Chatbots for Language Learning Across Disciplines
Saddleback College AI for Language Acquisition (AI FAST)
Chris Stillwell, Saddleback College
This poster shares lessons from a college-wide faculty initiative exploring how custom AI chatbots can support language learning and disciplinary literacy. Faculty teams designed and tested chatbots for English, Spanish, Italian, Arabic, Sociology, Business, and other subject areas, investigating how conversational AI can promote engagement, practice, and understanding. We highlight effective chatbot design principles, practical strategies for integrating chatbots into courses, and research on student perceptions of their impact on learning. The project also surfaced important insights about the opportunities and limitations of AI-mediated learning. Visit our poster to see examples of the chatbots in action, explore findings from the project, and access resources, websites, and tools available for adaptation and reuse.
Beyond Grades: Building Portable Student Domain Knowledge Models for Flexible Learning Pathways
Acquiring math prerequisites efficiently and scalably with AI (AI FAST)
Sesh Murthy, UC San Diego
As students move across courses, platforms, and credentials, much of what is known about their preparation is lost or reduced to grades. This work-in-progress explores whether student domain knowledge models can become reusable, portable infrastructure for flexible learning pathways. ASPIRE, an AI-supported math tutoring platform at UC San Diego, uses concept-level student models to capture what learners know, where they struggle, and what support they need next. By combining these models with generative AI, ASPIRE aims to provide more personalized guidance while giving instructors deeper insight into student understanding. We share early findings, key design challenges, and a vision for learner-centered records that support prerequisite recovery, degree progress, and workforce pathways. Challenges include measuring mastery with minimal student burden, representing learning more holistically, and enabling adoption across educational systems.
Reimagining the Traditional Calculus Sequence for Data Science
INCLUDE (Data Science)
Mikahl Banwarth-Kuhn, Cal State East Bay
What would mathematical preparation look like if we designed it for today’s students, rather than inherited it from yesterday’s disciplines? This poster describes the development of a new transfer-level mathematics pathway designed to rethink the role of calculus in undergraduate education and reimagine mathematical preparation for a world shaped by data, computation, modeling, and complex systems. Rather than emphasizing symbolic manipulation alone, students engage with modeling, rate of change, graphical reasoning, approximation, iteration, and computational exploration through authentic applications. We share the rationale behind the redesign, examples from the curriculum, early implementation experiences, and lessons learned in creating transfer-aligned pathways. Visit our poster to discuss how organizing mathematics around competencies rather than inherited course structures can expand access, maintain rigor, and better prepare students for diverse academic and career pathways.
Computer-Based Testing + Equitable Grading = A’s For All
A’s-for-All (A4A): Scaling Mastery Learning Through Technology, Advocacy, Policy, and Partnerships (Scaling Success)
Armando Fox, UC Berkeley
Dan Garcia, UC Berkeley
Bloom’s classic work on mastery learning showed that when students have multiple opportunities to practice and demonstrate mastery, achievement rises and equity gaps shrink. In principle, any student willing to do A-level work should be able to earn an A, without lowering standards. Scaling this approach, however, requires technology and automation to support repeated assessments, flexible policies, and sustainable course operations. This poster shares UC Berkeley’s efforts to make that vision practical. We describe the growth of a computer-based testing facility from one course with 230 students in 2023 to more than 16 courses serving 5,500 students in 2026. The platform supports mastery-learning practices such as second-chance assessments, “clobber” exams, flexible deadlines, and standards-based grading, while also strengthening assessment integrity in the age of AI. We highlight successes, surprises, and ongoing challenges from real-world implementation, offering practical insights for faculty interested in equity, student success, and evidence-based approaches to assessment.
Human Value that Grows with AI Capability
*Not a Learning Lab funded project
Sathya Narayanan and Michelle Skoor of the Computing Talent Initiative, Cal State Monterey Bay
As AI capabilities accelerate, postsecondary institutions face two questions: which human capabilities matter most, and how can we help diverse learners develop them? A core challenge is a structural mismatch: a fast, boundary-crossing technology meeting a slow, boundary-bound educational system. For more than 15 years, the Computing Talent Initiative (CTI) at Cal State Monterey Bay has worked at that intersection. This poster presents a framework grounded in Symbiotic Thinking, the human-led pursuit of wisdom in partnership with human and artificial intelligences. The framework develops Self-Directed Learning, Integrative Solving, and Adaptive Building toward two outcomes: Superagency and Human Value. Our pilot research has spanned high schools, colleges, and workplaces, and we welcome feedback, questions, and collaborators.
AI as Rescue, Not Shortcut: Productive Struggle in Gateway STEM Courses
AI-Enhanced Curriculum: Transforming Calculus and Statistics Education (AI FAST)
Sergio Hidalgo, Cal State Fullerton
As generative AI enters more classrooms, the key question is not whether students will use AI, but how course design can ensure it strengthens thinking rather than replace it. This poster shares results from Pasadena City College’s project integrating MathGPT.ai, ChatGPT, and Python-based modeling into Calculus and Statistics courses. Using an “Attempt-Chat-Reattempt” approach, students are guided to attempt challenging problems, then use AI feedback to identify errors, revise their reasoning, and try again. Data from more than 53,000 student problem attempts in pilot sections show that productive struggle before AI support led to stronger outcomes relative to comparison sections, with especially notable gains in course sucess for Pell-eligible students. These findings suggest how instructional design—not AI access alone—can determine whether generative AI advances reasoning, equity, and student success.
WITH SUPPORT FROM
