Factorized execution
Represent repeated data once while preserving the logical combinations required by a query.
RepresentationData systems researcher · Montréal
Building query engines that do less work with data.
I’m Sunny, a PhD candidate at Polytechnique Montréal. I combine factorization, vectorization, and hardware-conscious algorithms to make analytical and semantic queries faster and more memory-efficient.
Factorized execution
Vectorized processing
Compression-aware systems
Modern hardware
01 / About
From algorithms to working systems.
02 / Research
Making compressed data practical to compute on.
My work connects data representation, execution strategy, and hardware behavior—three layers that are too often optimized in isolation.
Vectorized processing
Process data in cache-friendly batches to turn compact intermediate results into efficient execution.
ExecutionHardware-conscious design
Shape algorithms around memory movement, cache behavior, and the realities of modern processors.
Architecture03 / Selected work
Publications
04 / Recognition
Honors & awards
Recognition for research potential and engineering impact.
Let’s connect