Partnered with Qualcomm's Mobile Chip Architecture team to build performance infrastructure and next generation of mobile profiling tools, aimed at gaining critical insights into hardware / software performance.

Focused on prototyping techniques to monitor power / performance and techniques to monitor user interface responsiveness issues; drove all into production tools.Developed utility (IPPET) to display CPU, graphics processing unit (GPU), battery, network, c-states and disk power usage for each process over time; early version was utilized by Fujitsu and Intel Apple teams, and was also shown to Microsoft for ease of monitoring power on recent Intel platforms.Created utility (ySoSlow) to monitor user experience responsiveness, power, and performance on Windows using ETW for near-zero-overhead background data collection; exposed responsiveness issue due to platform overheating by fast battery charging.Other roles in Intel include Computational Scientist assisting scientists optimizing code at National Labs using Intel Asci Red supercomputer. Also building tools to characterize, monitor and help optimize code on desktops and servers.