Ali Anwar

I am an Assistant Professor in the Computer Science and Engineering Department at the University of Minnesota, where I build systems that make machine learning and data-intensive computing faster, cheaper, and more trustworthy across the full stack, from storage and serverless runtimes up through distributed training and large-model inference. My core focus is efficient reasoning and systems for large language models: by storing, reusing, and pruning the intermediate computation that models would otherwise discard, my group cuts the cost and latency of inference without sacrificing answer quality. A second thread makes these systems trustworthy, spanning federated learning under privacy, robustness, and heterogeneity constraints, and AI safety and alignment through distributed architectures for more reliable ML. I also work on serverless and FaaS computing across the software–hardware spectrum, and on scheduling for distributed quantum computing that jointly accounts for compute, network, and data locality.

My work has received five Best Paper awards, multiple awards from IBM, the Samsung GRO award, and the Pratt Fellowship award, and I am a co-inventor on more than 15 US patents. Before joining UMN, I was a Research Staff Member on the AI Platforms team at IBM Research Almaden. I received my Ph.D. in Computer Science from Virginia Tech, advised by Dr. Ali R. Butt, and spent three summers at IBM Research during my doctorate. Earlier in my career I worked as a tools developer on several open-source projects, including GNU GDB, Embedded Linux, U-Boot, and BusyBox.

Recent News

Sponsors

We gratefully acknowledge the support of our sponsors: