The Columbia Center of AI Technology (CAIT) in collaboration with Amazon was founded in 2020 with a mission to better society through the development and adoption of advanced AI technologies contributing to a more secure, connected, creative, sustainable, healthy, and equitable humanity.
In keeping with CAIT's mission, our 2025 CAIT Spring Symposium and Research Showcase - the 5th annual event - will feature research highlights from past and current CAIT projects. We'll also look to AI's future with a panel on AI agents featuring Columbia and Amazon researchers.
Columbia students, stay tuned for more information about a concurrent info session about careers at Amazon!
Agenda
9:30 AM | Registration Opens
10:00 AM | Welcoming Remarks
Shih-Fu Chang, Dean, Columbia Engineering; Director, CAIT
10:10 AM | Keynote: GenAI and Automated Reasoning for Building Web Services
Serdar Tasiran, Principal Applied Scientist, Amazon Web Services
About the Keynote: I will describe three use case scenarios that users building on web services encounter, encouraging researchers to explore methods that combine automated reasoning (formal methods) for program analysis with GenAI to make humans more effective in these settings. These scenarios are (i) building applications that use web service APIs, (ii) using infrastructure as code to manage deployment of cloud resource configurations, and (iii) migrating without disruption from on-premises computing using legacy programming languages to computing on the cloud with modern programming languages. The talk will focus mainly on the application challenges and highlight the domain constraints that we believe motivate the use of automated reasoning in conjunction with GenAI.
10:55 AM | CAIT Research Talks
Using Speech and Language to Identify Patients at Risk for Hospitalizations and Emergency Department Visits in Homecare
Maryam Zolnoori, Columbia School of Nursing & Zoran Kostic, Electrical Engineering
EDITGUARDIAN: A Universal Multi-Agent LLM Framework for Semantic-Preserving Code Editing
Junfeng Yang, Computer Science
Towards Causal Artificial Intelligence
Elias Bareinboim, Computer Science
12:00 PM | Lunch & Networking
Carleton Commons
1:30 PM | Student Lightning Talks
Zach Horvitz, CAIT PhD Fellow
Kevin Xia, CAIT PhD Fellow
1:45 PM | CAIT Research Talks
Robotic Manipulation using Low-Cost Force/Torque Sensing
Zhanpeng He, Computer Science & Sharfin Islam, Mechanical Engineering
Presenting work performed with Matei Ciocarlie, Mechanical Engineering & Ioannis Kymissis, Electrical Engineering
Neural Methods for Analyzing and Interpreting Works of Art
Kathy McKeown & Carl Vondrick, Computer Science
Dataset Search for Data Tasks
Eugene Wu, Computer Science & Rachel Cummings, Industrial Engineering and Operations Research
2:45 PM | Break
3:00 PM | Panel on AI Agents
Yunzhu Li, Computer Science
Danielle Perszyk, Amazon SF AGI Lab
Alex Williams, AWS Bedrock
Zhou Yu, Computer Science
4:30 PM | Reception & Poster Session
Carleton Commons
Featuring posters from student researchers Zach Horvitz, Kevin Xia, William Ho, and Hantao Yu.
About the Speakers
Serdar Tasiran is a Principal Applied Scientist at Amazon Web Services. He received his PhD in EECS from UC Berkeley in 1998. He has led automated reasoning teams at Amazon since 2016. Prior to Amazon, he was a faculty member at Koc University, Istanbul, Turkey 2003-2016, and a research scientist at the Systems Research Center (DEC/HP/Compaq) 2000-2003. While at Koç University, he had visiting appointments at Microsoft Research, Univ. Paris VI, Univ. Paris VII, MIT, and EPFL. He was a Distinguished Research Fellow at Sabanci University, Istanbul, Turkey, and on the external advisory board for computer engineering at Koc University.
Danielle Perszyk is a cognitive scientist that leads the human data team at the Amazon SF AGI Lab, which is building a general-purpose agent. Prior to Amazon, she was at Adept and Google. She did her PhD at Northwestern, where she studied the evolutionary and developmental origins of human language.
Alex Williams is a senior applied scientist in the AWS Agentic AI organization where he works on problems related to human-computer interaction, synthetic data generation, and agent evaluation. Before joining Amazon, he was an assistant professor in the University of Tennessee, Knoxville's Electrical Engineering and Computer Science Department and a postdoctoral researcher in the University of California, Irvine's Informatics Department. He holds a Ph.D. in computer science from the University of Waterloo.
