Systems Engineer | AI Infrastructure | Columnar Data | B.Tech IT @ MIT Manipal
I build high-performance scientific and AI systems across Python C++, GPU acceleration, and columnar data layouts.
My work includes backend abstraction for large-scale array systems, zero-copy data interfaces, and GPU-based neural decoding on NVIDIA A100.
Active contributor to the Scikit-HEP ecosystem, primarily Awkward Array, improving performance and correctness for nested scientific data.
- GPU Pipelines: CUDA-backed backend abstractions
- Data Interoperability: Apache Arrow ↔ PODIO ↔ C++ ↔ Python
| Project | Technology |
|---|---|
| Awkward Array | C++ Python Apache Arrow cuDF |
| Brain-to-Text GPU Pipeline | Python PyTorch CUDA |
| Research: Knowledge Graph → Text (IEEE MoSICom '25) & Extended OJCS Submission | LLMs NLP Python |
| Arrow ↔ PODIO Interface | C++ Modern CMake Arrow |
