Mar 2023 – Dec 2023 · 9 mos

Mar 2019 – Jan 2023 · 3 yrs 10 mos · Greater Seattle Area

• Working on Artificial Intelligence/Machine learning Platform and Infrastructure. Focusing on the design, implementation and improvement of distributed ML big data reading and processing pipeline for machine learning training workloads. Recent projects include:
• service auto-scaling;
• service reliability and efficiency observability;
• service performance benchmarking framework, tuning and improvement for both CPU and GPU clusters.
• Worked on data storage infrastructure team for new virtual global storage pool auto-migration.
• Language: C++/Python
• Platform: Linux

Apr 2017 – Mar 2019 · 1 yr 11 mos · Greater Seattle Area

• Working on distributed system such as software load balancer cluster. Developing control plane that provides multiple functionalities such as L3 tunneling, L4 forwarding, L7 proxying, TLS termination, authentication, firewall, etc. Enhancing data plane performance from different network layers and system level.
• Using Kubernetes orchestration technology to speed up distributed system's evaluation and deployment (CI/CD). Adopting micro-services design philosophy to design and implement system availability, scalability and robustness.
• Language: Go/C/C++
• Platform: Linux

Mar 2013 – Apr 2017 · 4 yrs 1 mo · Wichita, Kansas

• Working on Linux-based RTOS for high performance storage array. Enhance system performance via multiple technologies such as bypass kernel, zero copy, etc. Design and implement user space memory management infrastructure, real-time task scheduler and tasks synchronization APIs.
• Working on EClouds - Cloud-based Data backup/recovery for Storage Array. Participate in the design of network security policies for storage array. Participate in the design and implementation of network functionality such as DHCP, DNS and NTP etc.
• Working on pulling Linux Infiniband driver to VxWorks environment.
• Language: C/C++
• Platform: Linux and VxWorks

May 2012 – Aug 2012 · 3 mos · Palo Alto

• Project: Performance Prediction of Migration module (vMotion) for ESXi Server.
• 1. Design and develop a framework to test and predict vMotion’s impact on virtual machines' performance.
• 2. Discovery more accurate metrics to measure guest performance from Hypervisor viewpoint.
• 3. Design algorithm to predict vMotion network traffic accurately.

Aug 2006 – Feb 2013 · 6 yrs 6 mos · Manhattan, KS

• Project: performance evaluation and improvement of peer-to-peer network.
• 0. Focus on a P2P file-sharing network adopting distribute hash table tech and having millions of users.
• 1. Design and implement distributed framework for measuring the performance of Distributed Hash Table based P2P network.
• 2. Measure and analyze 10,000+ nodes’ overlay routing tables.
• 3. Mitigate weakness of design by providing several solutions.
• 4. Make Publications on professional Journal and conference.