The AICE 2026 Fall Symposium served as a reminder of the importance of the collaboration between Amazon and Illinois as evidenced by Ruhi Sarikaya, AICE co-director, vice president Alexa AI and CS adjunct research professor, who, standing before a full room at the Thomas M. Siebel Center for Computer Science giving his welcome speech, pointed out how busy this time of year is at Amazon with their fall devices launch event.

“I said, ‘No, I’m going to Illinois. Put it on the calendar.'"

Sarikaya welcomed guests to the 2026 AICE Fall Symposium which took place on Sept. 10 at the Siebel Center and, by his count, this was the seventh meeting since Amazon began its relationship with the university in 2015 when the first Illinois proposal received Amazon research funding.

“I haven’t missed a single one,” Sarikaya said.

Photo Credit: Rudy San Miguel/Siebel School

The Amazon-Illinois Center on AI for Interactive Conversational Experiences (AICE) was launched in 2023 in the Siebel School of Computing and Data Science at The Grainger College of Engineering. At that time, AICE made its first call for proposals, resulting in the awarding of two fellowships and funding for five projects. That first year, Amazon had just over 35 Illinois Ph.D. interns. In 2026, that number had close to doubled.  Sarikaya said that Illinois consistently sends the most interns to Amazon.

Areas of collaboration include Alexa AI, AGI, Prime Video, AWS AI, AWS Automated Reasoning, and Amazon Health Science. Sarikaya shared future research directions that included advanced natural language understanding, enhanced contextual understanding, human-AI collaboration and novel applications including education and healthcare.

“I’m excited about healthcare,” Sarikaya said. “I think there will be some big disruptions and innovations coming up in the next couple years, like curing major diseases. I genuinely believe that.”

Photo Credit: Rudy San Miguel/Siebel School

The day would continue to highlight the impact of the 11-year collaboration.

Keynote speaker Manling Li, CS assistant professor at Northwestern University, shared her research, “Toward Controllable Foundation Agents,” which posits whether agents can be made more controllable and safe.

Yue Guo, iSchool assistant professor and CS affiliate professor, also provided a keynote address, “Medical AI Needs a Curriculum. Heng Ji, CS professor and AICE co-director, introduced Guo and pointed out that healthcare is a vital area for making AI useful and responsible.

“To have responsible AI for healthcare, you need to know how AI works, how responsible AI works and how medicine works,” Ji said. “Professor Guo is the leading expert to cover all three.”

Guo, who is also affiliate faculty with Siebel School, tackles her research with two important questions in mind: What can the technology really do and what do the physicians, clinicians and patients need from those systems in practice? In her keynote speech, she explored how we can train medical AI to use knowledge effectively with people, a matter Guo sees as urgent, given AI’s current presence in hospitals.

Keynote speaker Ashutosh Joshi, principal scientist with Amazon Health AI, shared his research on "Safety-First Evaluation of Conversational AI." Ismini Lourentzou's keynote speech centered on "Generative Models for Physical Interactions." Lourentzou, iSchool assistant professor and CS affiliate professor, noted that physical intelligence requires models that preserve relevant information, predict how the world responds to action and use sensory feedback to adapt behavior. Her research explores how connecting these capabilities through multimodal understanding and generation can help AI systems reason about physical and social interactions and act more reliably.

The AICE 2026-27 research projects were also shared:
•    Reyhan Jabbarvand, CS assistant professor, Siebel School, “Process-Centric Analysis for Robust Evaluation, Steering, and Training of AI Programming Agents”
•    Tal August, CS assistant professor, Siebel School, “Evaluating Dialogue-Based Personalization End-to-End with Human Users”
•    Jiaqi Ma, CS Assistant professor iSchool and CS affiliate professor, “A Data Regularization Framework for LLM Post Training”
•    Sagnik Mukherjee, CS Ph.D., presented on behalf of Dilek Hakanni-Tur, Siebel School CS professor and Amazon Scholar, “Scaling Multi-Turn Reinforcement Learning in LLMs”

A Fireside Chat followed, moderated by Ji, and focusing on “Training the Next AI Force in the LLM Era.” On the panel were Hakanni-Tur and Sarikaya, joined by Nancy Amato, Abel Bliss Professor and Siebel School director. 

The 2026-27 AICE Fellows were the final speakers of the day:
•    Haozhe Zhao, CS PhD, AICE Fellow, “A Multi-Agent Harness for Editable Scientific Figure Generation from Diverse Inputs”
•    Jinu Lee, CS PhD, AICE Fellow, “Structure-Aware Formal Verification of Reasing Traces”

The symposium closed out with a poster session featuring the research of current CS students. 

Grainger affiliations:

Ruhi Sarikaya is an Illinois Grainger adjunct professor of computer science in the Siebel School of Computing and Data Science. 

Yu Guo is an assistant professor with the School of Information Sciences and an affiliate professor of computer science in the Siebel School of Computing and Data Science. 

Heng Ji is an Illinois Grainger professor of computer science in the Siebel School of Computing and Data Science and an affiliated electrical and computer engineering faculty member in the Department of Electrical and Computer Engineering and the Coordinated Science Laboratory.

Ismini Lourentzou is an assistant professor with the School of Information Sciences and an affiliate professor of computer science in the Siebel School of Computing and Data Science. 

Reyhan Jabbarvand is an Illinois Grainger professor of computer science in the Siebel School of Computing and Data Science. 

Tal August is an Illinois Grainger professor of computer science in the Siebel School of Computing and Data Science. 

Jiaqi Mais an assistant professor with the School of Information Sciences and an affiliate professor of computer science in the Siebel School of Computing and Data Science. 

Dilek Hakkani-Tür is an Illinois Grainger professor of computer science in the Siebel School of Computing and Data Science. 

Nancy M. Amato is an Illinois Grainger professor of computer science and the Abel Bliss Professor of Engineering in the Siebel School of Computing and Data Science as well as director of the Siebel School.