Brown CS News

Google LLC, the multinational corporation and technology company, has just announced the Google ML and Systems Junior Faculty Awards to recognize machine learning and systems pioneers in academia, and three Brown CS faculty members are among the more than fifty inaugural recipients. Deepti Raghavan, Malte Schwarzkopf, and Nikos Vasilakis were chosen by a distinguished group of Google engineers and researchers for work that leads the analysis, design, and implementation of efficient, scalable, secure, and trustworthy computing systems. Their work crosses the technology stack, from algorithms to software and hardware, enabling machine learning and cloud computing at an increasingly massive scale. Each will receive grants of $100,000 in unrestricted funding.

Deepti’s research group focuses on reliable and efficient networked and machine learning systems. They develop new networked-systems abstractions that enable new opportunities for performance, particularly by modifying APIs that affect when, how, and where data is moved. Some recent systems include Alto (EuroMLSys’24), which introduces an API to enable streaming and parallelization in emerging AI applications, improving performance over existing solutions, and Cornflakes (SOSP’23), which improves serialization overhead in low-latency datacenter systems. Deepti also collaborates with the Brown Database Group on building reliable AI data processing systems (VLDB’25).

Malte’s Efficient and Trustworthy Operating Systems (ETOS) research group builds and investigates core abstractions and techniques for improving the efficiency and trustworthiness of practical computer systems. The ETOS group conducts research on operating systems, privacy and security, and data center systems, and has a long track record of building practical, efficient, and easy-to-use working systems released as open-source software. Some recent systems include Paralegal (OSDI’25), Sesame (SOSP’24), and K9db (OSDI’23), which help developers avoid privacy-related bugs in their software and provide technical means to ensure that organizations manage data responsibly; as well as Mach (SOSP’25, to appear), Quicksand (NSDI’25), Nu (NSDI’23), and AIFM (OSDI’20), which improve the performance of datacenter systems software for observability, resource-efficient programming, and remote memory by orders of magnitude over state-of-the-art systems.

Nikos’ Analysis and Transformation in Languages, Applications, and Systems (ATLAS) research group builds systems that lift the capabilities of developers dealing with the vast complexity of modern software systems by automating away inessential complexity and automating in desired features — for example, automatically scaling out or securing systems that use many software dependencies. Recent ATLAS systems such as Fractal (NSDI’26), Slowpoke (NSDI’26), and DiSh (NSDI’23) offer order-of-magnitude performance improvements automatically, BinWrap (AsiaCCS’26) automatically protects applications against supply-chain insecurity, Sash (HotOS’25) ensures the correctness of complex polyglot systems, and standardized collections of benchmarks such as Koala (ATC’25) and SecBench.js (ICSE’23) enable further scientific research in these and other domains. ATLAS research is available as open-source systems — for example, ones available by the Linux Foundation, ones underpinning cybersecurity offerings, and ones contributed to by dozens of developers.

All three groups are part of Systems@Brown, the Brown University Systems Research Group, which investigates questions in computer systems, particularly distributed systems, operating systems, networking, and security. They build practical systems that combine performance, correctness, automation, and other desirable properties.

For more information, click the link that follows to contact Brown CS Communications Manager Jesse C. Polhemus.