My research interests are in systems support for data science and big data to enable efficient use of modern and future hardware. My research spans across multiple system subfields: data processing layer, operating systems, and hardware accelerators for data processing.
I am joining the Department of Computing at Imperial College London as a Lecturer (Assistant Professor) starting from October, 1st 2017.
Thesis: Database/Operating System Co-Design
Adviser: Prof. Gustavo Alonso, Co-adviser: Prof. Timothy Roscoe
FPGA Based Data Partitioning.
ACM SIGMOD 2017, To appear
BatchDB: Efficient Isolated Execution of Hybrid OLTP + OLAP Workloads.
ACM SIGMOD 2017, To appear
Customized OS support for data-processing .
In Proceedings of the 12th International Workshop on Data Management on New Hardware (ACM DaMoN) 2016
Broom: sweeping out Garbage Collection from Big Data systems.
In Proceedings of the 15th USENIX Conference on Hot Topics in Operating Systems (HotOS) 2015
Deployment of Query Plans on Multicores.
In Proceedings of the VLDB Endowment, Volumne 8, Issue 3, 2014
COD: Database/Operating System co-design.
Online proceedings of the 6th Biennial Conference on Innovative Data Systems Research (CIDR) 2013
Towards Database/Operating System co-design.
Second Workshop on Systems for Future Multicore Architectures (SFMA) 2012
Database Engines on Multicores, Why Parallelize When You Can Distribute?.
In Proceedings of the 6th Conference on Computer Systems (EuroSys) 2011
Co-chair: SIGMOD SRC (Student Research Competition) 2018
PC member: SIGMOD Demo 2017, ISCA's ASBD (Architecture Support for Big Data) 2017
Shadow PC: EuroSys 2016
Organizer: "Panel for Women in CS" at ETH Zurich (Spring 2015, 2016)
Google PhD Fellowship for Operating Systems - 2014
Best Poster Award at EuroSys - 2012
Worked on better automatic memory management in runtime engines
for the Naiad big data system. Contributed to the system design,
and worked on automatic inferral of typed regions.
Supervised by: Derek G. Murray and Michael Isard.
Explored partition-reduce SQL operators, as a common
method for enhancing existing SQL operators with hardware tuned
partitioning primitives, and its applicability for Oracle's
Supervised by: Michael Duller, Garreth Swart, and Nipun Agrawal.
Internship as part of the MICS internship program. Explored and
evaluated a data processing architecture on multicore machines for
a mix of workloads. Results were published at EuroSys.
Supervised by: Tudor-Ioan Salomie and Gustavo Alonso.
Analysis of rhetorical aspects in mathematical documents. Implemented
appropriate mark-up to support the new elements and a visualization
Supervised by: Christoph Lange and Tudor Groza.