The Jetstream2 AI Fellows Program is an exciting new initiative that provides a unique opportunity for both early-career and established researchers to elevate their AI-based research to the next level. Set to run from January through September 2025, this nine-month program aims to equip participants with the expertise and knowledge necessary to harness the full potential of Jetstream2’s state-of-the-art resources. By focusing on practical application and community impact, the program offers a comprehensive pathway for researchers to achieve significant advancements in their fields.
Built around four key themes, the Jetstream2 AI Fellows Program is specifically designed to support participants at various stages of their research journey. The first part of the program offers fellows personalized consulting services to help them navigate the technical and logistical challenges, including onboarding assistance, as well as expert guidance on how to make the most of Jetstream2’s advanced GPU resources.
“The AI software technology stack is rapidly evolving and comprises multiple components that are often challenging to install and configure but they are essential for effectively managing AI workloads. The AI fellows will learn the best practices of utilizing these components e.g. MLFlow, PyTorch, DVC, vLLM etc. and customize those for their project, creating a reproducible community resource that can be readily deployed and utilized by others,” stated Nirav Merchant, co-Principal Investigator of the Jetstream2 cloud system.
In addition to research project consulting, the program also provides comprehensive training through the Jetstream2 AI Workbench as well as other targeted sessions designed to augment participants’ AI techniques. A major highlight is that the Jetstream2 program offers fellows priority access to powerful, high-end GPUs, which include 90 GPU-enabled nodes featuring NVIDIA A100 GPUs, with new NVIDIA H100 GPUs arriving in the coming weeks. Whether it be refining their existing skills or exploring new approaches, the training and access offered to researchers selected for the program can help participants further their research, be it complex computations or methodical analyses at a faster pace through advanced AI techniques.
“We are very excited to offer these new, powerful GPU resources and consulting support to the science and engineering research community. Democratizing access to these scarce resources has always been a top priority for Jetstream2,” said David Y. Hancock, Principal Investigator of the Jetstream2 cloud system.
Beyond advancing individual research, the program is also designed to ensure that the knowledge and insights gained by fellows are shared with the broader research community. Fellows will collaborate with the Jetstream2 team to identify meaningful ways for the fellows to propagate what they have learned to others in their field through tutorials or other educational resources.
While currently limited to U.S.-based researchers and educators, graduate students are also encouraged to apply with a letter of collaboration from their graduate advisor. Apart from being judged on the technical feasibility of the project and its match with available Jetstream2 resources, the potential for the project to produce meaningful outcomes and contributions to the field will be a key driver in the selection process. Upon successfully completing the program, fellows will receive a $5,000 stipend and will be featured on one of Indiana University’s websites. The published feature will highlight their contributions and achievements, providing them with additional visibility within the research community.
“The Jetstream2 AI Fellows program is a wonderful opportunity for researchers to adopt new technologies and techniques. We are thrilled to be able to make these resources available to researchers from all domains of science and scholarship,” said Winona Snapp-Childs, PhD, Director of the Jetstream2 AI Fellows program.
Researchers and graduate students interested in applying can learn more and apply here by October 30.