PhD Alums Mohapatra and Cruz Selected for MLCommons Rising Stars

Northwestern Engineering’s Payal Mohapatra (PhD ’26) and Stefany Cruz (PhD ’25) attended the MLCommons Rising Stars Workshop, hosted last month at Advanced Micro Devices headquarters in Santa Clara, California.

Mohapatra and Cruz were among 39 outstanding early-career researchers from 26 institutions worldwide selected for the fourth annual MLCommons Rising Stars cohort. Demonstrating exceptional promise in machine learning (ML), systems, and data systems research, the cohort stands out for their current achievements and for their potential to shape the future of the field.

Launched in 2023 by MLCommons, an AI engineering consortium, the Rising Stars program supports and connects early-to-late-stage and recently graduated PhD students working at the intersection of ML and systems. Through this initiative, participants engage with a vibrant global community, connect with leaders across academia and industry, and further develop their technical and professional skills.

During the MLCommons Rising Stars Workshop, participants presented their research, explored emerging opportunities, participated in career development sessions, and built connections with peers and mentors across sectors.

“What I appreciated most was that the discussions did not stop at technical challenges. We talked just as much about societal impact, from the environmental toll of the ever-increasing power demands of data centers to how the AI wave is reshaping education, especially for computer science students,” said Mohapatra, a PhD alum in computer engineering who was advised by Professor Qi Zhu in the Design Automation of Intelligent Systems Lab. “We had wonderfully candid group discussions about the field we are inheriting and the one we want to build.”

From Best Computer Engineering Thesis to Rising Star

At Northwestern, Mohapatra and Cruz both won Best Computer Engineering PhD Thesis Awards from the Department of Electrical and Computer Engineering. Using methods inspired by machine learning, signal processing, and embedded systems, Mohapatra designs algorithms to extract more latent information about human behavior from multimodal time-series data from sensing applications like smartwatches, smart glasses, and fitness trackers. Mohapatra recently joined AI compute company Arm as a staff research scientist, where she is studying how intelligent systems should allocate limited resources—such as computation, sensing, memory, bandwidth, or energy—as they operate under uncertainty.

“The simple collect-data, train-model, deploy era of AI is giving way to systems that must adapt continuously, using feedback from their models, users, environments, and hardware to decide what information to gather and what computation to spend,” Mohapatra said. “I want to build systems that work under real-world resource constraints, with careful attention to energy, latency, and practical deployment.”

At the McCormick School of Engineering, Cruz was coadvised by Professor Maia Jacobs and Josiah Hester, adjunct associate professor. Cruz is currently a Washington Research Foundation Postdoctoral Fellow at the University of Washington’s Paul G. Allen School of Computer Science and Engineering. In her research, she builds on-device agentic AI for urban safety and sustainability, focusing on real-time, privacy-preserving sensing and decision-making.

Finding connection and inspiration

At the MLCommons Rising Stars Workshop, Mohapatra was motivated by advice from multiple speakers. She also met several like-minded researchers who could be potential collaborators on upcoming projects, and the workshop's emphasis on cross-layer thinking has already changed how she frames her research problems. The conversations at the workshop helped renew her commitment to building methods that are upfront about the full-stack costs rather than optimizing performance metrics in isolation.

“The energy strain of AI is real, and while those of us on the algorithms side tend to focus on reducing compute, the real bottleneck is often data movement,” Mohapatra said. “Both make systems thinking essential for ML researchers.”