June 3, 2026 · New York City
Trustworthy AI for Code Industry Roundtable
A curated gathering of leaders from industry and academia for a day of invited talks, sharp discussion, and agenda-setting conversations in New York City.
Trustworthy AI for Code
Columbia University × National University of Singapore
Hosted in NYC by
Local hosts welcoming the roundtable to New York City
About the roundtable
What this roundtable aims to produce
The goal is to develop a shared understanding of trustworthy AI for code, including open technical challenges, evaluation criteria, and concrete opportunities for collaboration across research and practice.
Shared agenda
Identify the most important unsolved problems and align on high-impact research directions.
Evaluation framework
Discuss meaningful criteria for trustworthiness, usefulness, and deployment readiness.
Community building
Lay the groundwork for an ongoing forum and potential joint publication on this topic.
Leadership
Organizers
The roundtable is co-organized by researchers at Columbia University, the National University of Singapore, and IBM Research Lab NYC with deep expertise in software engineering and AI reliability.
Abhik Roychoudhury
Professor, National University of Singapore
School of Computing
Baishakhi Ray
Associate Professor, Columbia University
Department of Computer Science
On the ground in NYC
Local Organizer
Eugene Wu
Associate Professor, Columbia University
DAP Lab, Department of Computer Science
Maja Vukovic
IBM Fellow
IBM Research, T.J. Watson Research Center
Xuan Liu
Director, AI for Code
IBM Research, T.J. Watson Research Center
Invited speakers
Keynote speakers
We are honored to have the following keynote speakers confirmed for the roundtable.
Maja Vukovic
IBM Fellow
IBM Research, T.J. Watson Research Center
Xuan Liu
Director, AI for Code
IBM Research, T.J. Watson Research Center
Franjo Ivancic
Engineering Manager, Google
Aditya Kini
Senior Staff Software Engineer, Google
Elena Glassman
Assistant Professor, Harvard University
John A. Paulson School of Engineering and Applied Sciences
Ofir Press
Princeton University
Princeton Language and Intelligence
Koushik Sen
Professor, University of California, Berkeley
Computer Sciences Division
Panel Discussion
Panel Members
We are honored to have the following members for our panel discussion.
Petros Maniatis
Senior Staff Research Scientist, Google DeepMind
Bogdan Vasilescu
Associate Professor, CMU
Rajdeep Mukherjee
Senior Applied Scientist, AWS
Eno Reyes
CTO & Co-founder, Factory
Junfeng Yang
Professor, Columbia University
Discussion Topics
- Trust: How can trust in agents be engendered?
- Working with agents: How does it compare with traditional software evolution?
Program
Agenda
A structured day of talks, discussion, and synthesis designed to move from perspective sharing toward a concrete collaborative agenda.
Morning
- 08:30Registration
- 09:00Welcome and opening remarks
- 09:15Keynote: Maja Vukovic and Xuan Liu (IBM)
- 09:45Keynote: Franjo Ivancic and Aditya Kini (Google)
- 10:30Keynote: Elena Glassman (Harvard) // From Efficiency to Understanding: The Speed-Learning Tradeoff in Human-AI Co‑Creation of Software
- 11:00Panel: Agents in the workspace — pros, cons, and what comes next
Midday
- 12:00Lunch
- 13:00Keynote: Ofir Press (Princeton)
Koushik Sen (UC Berkeley) // KISS Sorcar: A Stupidly-Simple General-Purpose and Software Engineering AI Assistant - 14:00Breakout roundtable discussions
Afternoon
- 15:30Tea break
- 16:00Summary presentations
- 17:00Closing
Discussion themes
Roundtable topics
Four themes will frame the day's talks and discussions, spanning skills, processes, human–agent dynamics, model training, and verification.
01
Future of software engineering
What new skills will engineers need, and how should we prepare the next generation for an AI-augmented profession?
02
New SE processes and workflows
Rethinking CI/CD pipelines and technical debt management in a world where AI agents author and modify code at scale.
03
Training models with a trust mindset
Pre- and post-training techniques that build reliability, calibration, and verifiable behavior into code-generation models.
04
The role of V&V
Verification and validation of AI-generated code, including AI-based V&V techniques for evaluating AI-generated artifacts.
Discussion Groups
Group 1: Future of Software Engineering
- Aritra Sengupta
- Bogdan Vasilescu
- Denys Poshyvanyk
- Elena Glassman
- Gabriel Ryan
- Jürgen Cito
- Mark Santolucito
- Martin Kellogg
- Rekha Singhal
- Saurabh Sinha
- Sungmin Kang
Group 2: New SE processes and workflows
- Daniel Rodriguez-Cardenas
- Eno Reyes
- Foutse Khomh
- Ofir Press
- Pramod Pratap
- Rahul Krishna
- Saurabh Jha
- Simin Chen
- Xiaofei (Max) Ma
- Yangruibo (Robin) Ding
Group 3: Training models with a trust mindset
- Aditya Kini
- Baishakhi Ray
- Genta Winata
- Jatin Ganghotra
- Junfeng Yang
- Koushik Sen
- Maja Vukovic
- Saikat Chakraborty
- Sanjay Arora
- Saurabh Pujar
- Shubham Ugare
Group 4: The role of V&V
- Abhik Roychoudhury
- Chengpeng Wang
- Franjo Ivancic
- Jeffrey Levenberg
- Martin Hirzel
- Petros Maniatis
- Rajdeep Mukherjee
- Rajeev Alur
- Raju Pavuluri
- Rangeet Pan
- Tianyi Zhang
- Xuan Liu
Community
Participating organizations
Participants come from leading universities and technology companies working on AI for software engineering.
Academic organizations
Industry organizations
Attendance
Confirmed participants
This list will be updated as confirmations are finalized.
Abhik Roychoudhury
National University of Singapore
Aditya Kini
Google
Aritra Sengupta
Amazon
Baishakhi Ray
Columbia
Bogdan Vasilescu
CMU
Chengpeng Wang
Purdue
Daniel Rodriguez-Cardenas
William & Mary
Denys Poshyvanyk
William & Mary
Elena Glassman
Harvard
Eno Reyes
Factory
Foutse Khomh
École Polytechnique de Montréal
Franjo Ivancic
Google
Gabriel Ryan
Microsoft
Genta Winata
Capital One
Jatin Ganghotra
IBM
Jeffrey Levenberg
Infosys
Jürgen Cito
TU Wien
Junfeng Yang
Columbia
Koushik Sen
UC Berkeley
Maja Vukovic
IBM
Mark Santolucito
Barnard
Martin Hirzel
IBM
Martin Kellogg
NJIT
Ofir Press
Princeton
Petros Maniatis
Google
Pramod Pratap
Infosys
Rahul Krishna
IBM
Rajdeep Mukherjee
Amazon
Rajeev Alur
University of Pennsylvania
Raju Pavuluri
IBM
Rangeet Pan
IBM
Rekha Singhal
Tata Consultancy Services
Saikat Chakraborty
Microsoft
Sanjay Arora
Red Hat Research
Shubham Ugare
Meta
Saurabh Jha
IBM
Saurabh Pujar
IBM
Saurabh Sinha
IBM
Shyam Ramji
IBM
Simin Chen
Columbia Unversity
Sungmin Kang
National University of Singapore
Tianyi Zhang
Purdue
Xuan Liu
IBM
Xiaofei Ma
AWS
Yangruibo Ding
UCLA
The first edition of this roundtable, held in Singapore in January 2026, brought together more than 30 leaders from industry and academia to discuss trustworthy AI for code.
Sponsored by