Rashmi Vinayak

Rashmi Vinayak is an Associate Professor in the Computer Science Department at Carnegie Mellon University, with a courtesy appointment in the Electrical and Computer Engineering Department. She is a member of both the Systems group and Theory group at CMU and leads TheSys research group. She is also affiliated with the Parallel Data Lab (PDL).

Her educational background includes a Ph.D. from UC Berkeley in 2016, followed by postdoctoral studies at the same institution.

Rashmi's research spans the intersection of computer/networked systems and information/coding theory. Her current focus is on robustness and resource efficiency in data systems across storage, communication, and computation. Key thrusts include storage systems, caching systems, and systems for machine learning. Her work on SIEVE, a cache eviction algorithm, has been widely adopted by industry including VMware, Google, Redpanda, and numerous open source libraries.

Her recent publications demonstrate a strong trend toward practical systems research with theoretical foundations, particularly in caching algorithms, storage systems, and machine learning infrastructure. Many of her papers have received best paper awards and industry adoption.

Notable awards include:

• Sloan Research Fellowship (2023)
• IEEE Information Theory Society Goldsmith Lecturer (2023)
• NSF CAREER Award (2020)
• Multiple USENIX NSDI Community (Best Paper) Awards
• VMware Systems Research Award (2021)
• Facebook and Google Research Awards

Rashmi has supervised numerous PhD, Master's, and undergraduate students, many of whom have gone on to prestigious positions at Harvard, Google, Meta, and other leading institutions. Her research has been generously funded by NSF, Sloan Foundation, Open Compute Project, Google, Facebook/Meta, VMware, and Amazon Web Services. She actively collaborates with industry partners including Google, Microsoft, NetApp, Facebook, Cisco, Intel and Cloudera.

She leads TheSys research group which focuses on designing next-generation data systems that are robust, efficient, and performant. The group takes a multi-disciplinary approach spanning computer systems, information theory, and machine learning.