Ian Foster is Senior Scientist and Distinguished Fellow, and director of the Data Science and Learning Division, at Argonne National Laboratory, and the Arthur Holly Compton Distinguished Service Professor of Computer Science at the University of Chicago. He has a BSc degree from the University of Canterbury, New Zealand, and a PhD from Imperial College, United Kingdom, both in computer science. His research is in distributed, parallel, and data-intensive computing technologies, and their applications to scientific problems. The Globus software that he co-designed is widely used as scientific infrastructures. He is convinced that intelligent agents are poised to transform both science and scientific infrastructure.
Presentation Title:
ModCon Core: A Framework for Persistent, Observable, and Collaborative AI Agents
Presentation Abstract:
Large language models have made it easy to build impressive demonstrations of AI agents but deploying agents that can operate reliably over long periods, collaborate with one another, and integrate with scientific and enterprise computing remains a significant challenge. The ModCon Core Agentic Framework addresses this gap by providing a modular foundation for building persistent, stateful, and observable AI agents.
This talk introduces the design principles and architecture of ModCon Core, including its support for agent lifecycle management, memory, tool integration, multi-agent collaboration, and execution across heterogeneous computing environments. Particular attention is given to the challenges of long-running scientific and operational workflows, where agents must maintain context, recover from failures, reason over evolving information, and coordinate with both people and other agents.
Drawing on applications in scientific discovery, high-performance computing, and autonomous laboratories, the talk will illustrate how ModCon Core enables the construction of robust agentic systems that extend well beyond simple chat interfaces. The presentation concludes with lessons learned, current capabilities, and a roadmap toward scalable, trustworthy agentic infrastructure for next-generation scientific and engineering applications.