I'm a Sr. Software Engineer specializing in building scalable software solutions, resilient cloud architectures, and leveraging distributed computing to solve hard problems.
Hello! I am a passionate Sr. Software Engineer focused on crafting high-performance, robust, and cleanly abstracted solutions.
I recently work extensively at Enphase Energy bridging real-time energy management needs with modern web frameworks. I've also poured myself into academic research comparing Apache Spark with MPI C++ to push the limits of cluster computing logic across astrophysics algorithms.
GPU-accelerated 3D reconstruction pipeline using Facebook's Fast3R model with CUDA, processing multi-view images into dense point clouds. Features real-time WebSocket streaming of point cloud data with raw Float32Array buffers, bypassing React state for high-performance rendering of 1M+ points. Containerized full stack with Docker Compose and NVIDIA Container Toolkit.
GPU-accelerated real-time object detection system using YOLO26 with CUDA for low-latency video inference. Engineered an interactive Streamlit dashboard featuring dynamic FPS monitoring, custom glassmorphic styling, and threshold sliders.
Designed a cloud-augmented MPI support library to reduce network latency and improve performance of distributed computations in virtualized environments. Implemented optimized communication layers over traditional MPI to handle high-latency cloud infrastructure efficiently.
Developed machine learning models for malware detection on given static and dynamic analysis datasets of executable programs utilizing Random Forest classification at IIT Kanpur.
M.Tech Thesis at IIT Kanpur comparing traditional MPI C++ with Apache Spark for scientific data analysis. Executed benchmark experiments on CPU, memory, and I/O intensive applications from astrophysics, meteorology, and combustion across multi-node cluster environments.