Experience

Jump to

Research Labs

Synergy Lab, Virginia Tech

Graduate Research Assistant

Blacksburg, Virginia

High-performance systems for scalable genomic analysis, medical imaging, and edge computing.

  • Studied SparkLeBLAST scalability on A64FX systems and worked with RIKEN collaborators on data layout, vectorization, and format-conversion bottlenecks.
  • Explored Apache Arrow and Arrow Flight for genomic analysis pipelines; contributed to low-dose chest CT enhancement work using 3D-DDNet and VGG-based models.

InfoVis Lab, Virginia Tech

Undergraduate Research Assistant

Blacksburg, Virginia

Interactive visual analytics for image sorting and human-in-the-loop machine learning.

  • Worked on Andromeda features for exploratory sorting of edamame seed images.
  • Built tools for clustering images using domain feedback, pretrained image projections, and tuned feature-space distance weights.

Crawford Lab, Virginia Tech

Undergraduate Research Assistant

Blacksburg, Virginia

Numerical methods and machine learning for real-time coupled-cluster calculations.

  • Worked on localized orbital spaces, mixed-precision arithmetic, GPU utilization, and machine learning for real-time coupled-cluster calculations.
  • Used the Pade approximant method to reduce simulation time for accurate results from about 2000 atomic units to about 200; developed signal-analysis and LSTM forecasting routines.

Internships

Oak Ridge National Laboratory

Student Intern

Oak Ridge, Tennessee

GPU kernel optimization for fused matrix operations on modern Nvidia architectures.

  • Built tiled GEMM kernels in CUDA and a profiling harness for GPU memory hierarchy, occupancy, and warp scheduling experiments.
  • Used Triton to implement and tune fused GEMM+SiLU kernels, reaching roughly 1.2-1.35x single-precision speedups over a PyTorch baseline on Ampere and Hopper GPUs.
  • Prototyped a full SwiGLU fusion in CUTLASS with a Python-callable C++ kernel interface.

RIKEN Center for Computational Science

Student Research Intern

Tokyo, Japan

Scalable search algorithms for identifying cancer-associated mutation combinations.

  • Worked on algorithms for identifying multi-hit genetic mutation combinations associated with specific cancer types.
  • Implemented a sparsity-aware weighted set-cover approach to reduce the mutation-combination search space and prepare for large runs on Fugaku.

Amazon Web Services

Software Development Engineer Intern

Boston, Massachusetts

Backend infrastructure for IoT indexing, progress tracking, and operational reporting.

  • Worked on indexing and querying workflows for millions of IoT devices across multiple databases.
  • Built a tracking feature for indexing progress, wrote status data into Elasticsearch, and used AWS CloudWatch for backfill metrics.