Primary Convener:
Iksha Gurung, University of Alabama in Huntsville

Convener:
Campbell Watson, IBM Research USA
Manikandan Padmanaban, IBM Research India
Jagabondhu Hazra, IBM Research India
Venkatesh Kolluru, University of South Dakota

Early Career Convener:
Devyani Lambhate, IBM Research

Chair:
Campbell Watson, IBM Research USA
Manikandan Padmanaban, IBM Research India

Scientific foundation models (FMs) are reshaping how complex problems are solved across disciplines. To fully realize their potential, it’s vital to examine the entire pipeline from data collection to real-world integration. This session invites contributions that cover the end-to-end process, including data acquisition, preparation, model deployment, and post-processing. We are interested in the downstream use cases where FMs have been fine-tuned for specific tasks, such as applications for Earth and space sciences. Presentations should address challenges and strategies in making data model-ready, ensuring deployment demands like speed, scalability, and accessibility. Additionally, we encourage discussions on how the results are interpreted and presented to end users, ensuring they are both actionable and reliable. We also welcome discussions on ongoing model management, including monitoring, retraining, and maintaining performance over time. This session will serve as a platform to share operational strategies, lessons learned, and best practices from real-world use of FMs.

Index Terms
1640 Remote sensing
1942 Machine learning
1960 Portals and user interfaces
1968 Scientific reasoning|inference

Suggested Itineraries:
Machine Learning and AI
Open Science and Open Data

Neighborhoods:
1. Science Nexus