CV dblp sofiyase@buffalo.edu

I'm a final year Ph.D. student in the Computer Science and Engineering Department at the University at Buffalo. My research interests are in developing robust and reliable mobile, cyber-physical, and mixed criticality systems, particularly for robotics and computer vision applications.

In the summer of 2019, I interned at Google, where I developed a machine learning model to inform prefetching decisions on Chrome to decrease in user-perceived page load latency. Before starting the Ph.D. program, I was a software engineer at edX for a year, where I worked on all the layers of the "full stack" - from creating an entirely accessible essay-writing module in Javascript for keyboard-only users, to countless management commands and APIs in Python and Django, to running and analyzing A/B tests with Optimizely, to devops and system administration for VEDA, edX's video encoding and delivery pipeline.

I graduated from Brandeis University with a degree in computer science in 2017.

Publications

Sofiya Semenova, Steven Ko, Yu David Liu, Lukasz Ziarek, and Karthik Dantu. A Comprehensive Study of Systems Challenges in Visual Simultaneous Localization and Mapping Systems. ACM Transactions on Embedded Computing Systems (TECS), August 2024.

Ali J. Ben Ali, Marziye Kouroshli, Sofiya Semenova, Zakieh Sadat Hashemifar, Steven Y. Ko, and Karthik Dantu. Edge-SLAM: Edge-Assisted Visual Simultaneous Localization and Mapping. ACM Transactions on Embedded Computing Systems (TECS), August 2022.

Sofiya Semenova, Steven Y. Ko, Yu David Liu, Lukasz Ziarek, and Karthik Dantu. A Quantitative Analysis of System Bottlenecks in Visual SLAM. HotMobile 2022. Best Poster Award

Ali J. Ben Ali, Sofiya Semenova, and Karthik Dantu. Platform Variability in Edge-Cloud Vision Systems. HotMobile 2019.

Ryan Marcus, Olga Papaemmanouil, Sofiya Semenova, and Solomon Garber. NashDB: An Economic Approach to Fragmentation, Replication and Provisioning for Elastic Databases. SIGMOD 2018.

Ryan Marcus, Sofiya Semenova, and Olga Papaemmanouil. A Learning-based Service for Cost and Performance Management of Cloud Databases (Demonstration). ICDE 2017.

Projects

Anodize: Language-Level Guarantees for Mixed Criticality Systems

Real-Time Video Question Answering

Visual Question Answering survey

Edge-Assisted Trackerless Augmented Reality

Brandeis senior thesis: “Clouds That Think: Applications of Machine Learning Techniques for Elastic Databases".

For older projects, see the projects page of this website.