Working on Apple Intelligence and other on device ML challenges in Apple. Leading projects in benchmarking ML for current and future hardware. Implementing performance tooling support for internal and external performance tools.

Senior Software Engineer in CoreML Power and Performance Profiling Group.* Working as a Technical Lead on multiple future hardware performance analysis tools targeting ML inference.* Participating in bringing up and analyzing performance issues for new SOCs* A large amount of work that I have been driving for the past few years in the MLPT power and perf team has finally made it to external developers in the CoreML Instrument and the XCode Performance Report

Engineer in Radeon GPU Profiler team. My main role in the team focuses on processing low level GPU performance data for high level tools to provide actionable steps for performance optimizations.Previously, I lead the development of AMD's ROCm-GDB product which delivers hardware assisted debugger functionality to application developers. This project included close collaboration with compiler teams on debug information generation and user mode runtime teams. The debugger also included an Open source SDK for external tools developers.I also contribute in the other developer tools for GPUs and APUs at AMD.

Working on the development of debugging and profiling toolchains

Performing Computer Architecture Related ResearchWorking on Performance Analysis of General Purpose Computation Using Graphics Processing Units (GPU).Studying different programming models of NVIDIA (CUDA) and AMD (Brook+) for GPU computation

Working on OpenCl applications and tools as part of the Shader Compiler Group

Worked on implementing the Statistical Functions in MATLAB's Statistics Toolbox using Interactive Supercomputing's proprietary Star-P Platform.Wrote the Probability Distributions, measures of central tendencies, data clustering functions.Also analyzed corner cases and carried out regression tests.Was involved in writing and studying MPI implementation for some elementary data clustering algorithms.

Co-op Performance Engineer