html templates

Nosayba El-Sayed

Mobirise

I am a Lecture-Track Faculty in the Computer Science department at Emory University.

Bio.  Before joining Emory, I was a Postdoctoral Researcher at MIT, working on joint projects with Qatar Computing Research Institute (QCRI) in the area of data-driven system optimization in modern datacenters. I worked with Dr. Daniel Sanchez at MIT and Dr. Xiaosong Ma at QCRI. Before that, I completed my PhD at the University of Toronto with Dr. Bianca Schroeder. My thesis focused on the analysis and design of reliable datacenters. I also interned at Amazon Web Services where I worked on the prediction of power-outages in Amazon's datacenters.

Contact information available at Emory's CS Website.

Recent Stuff 
○ Emory ProgramHers club are having an exciting Grace Hopper Conference Panel, featuring students who have been to the conference before, to talk about the 2020 conference scholarship opportunities. Check it out!

○ I will be presenting our accepted Supercomputing'19 paper "Spread-n-share: improving application performance and cluster throughput with resource-aware job placement" during Supercomputing'19 in Denver, CO. This work is in collaboration with Tsinghua University and QCRI.

○ I will be joining my colleagues Dr. Vaidy Sunderam and Dr. Dorian Arnold as representatives of the CS dept. at Emory, in an NSF workshop focusing on designing departmental plans for Broadening Participation in Computing (BPC).  The workshop will take place in University of illinois, Urbana-Champaign (UIUC). 

Teaching Activities
○  Spring 2020: I will be teaching
     ○  CS171 - Introduction to Computer Science II
     ○  CS377 - Introduction to Database Systems

Research Activities
Large-scale systems, like supercomputers and cloud-computing datacenters, are quite challenging to manage (trust me, I wrote a thesis about it).  I am generally interested in designing and implementing data-driven techniques that exploit the different kinds of logs generated in these systems to improve how they operate, while utilizing simple machine-learning techniques. See my research page for more details on past and current projects.