(Jim) Zhiyuan Li

software engineer | research devoter | AI/ML enthusiast

CV

ABOUT ME

I am Zhiyuan Li and my nickname is Jim. Welcome to my personal website.


I am currently a software engineer at Amazon 126 Lab working on intelligent speaker device Echo and e-reader Kindle. Before that, I was a R&D Engineer at Cadence Design Systems working on graph parititon in VLSI and datastream infrastructure.


I am highly interested in research topics pertinent to traffic system optimization, machine learning, graph & system analysis and software development. In 2020, while working at Cadence, I am fortunate to be advised by Professor Charles (Chuck) Eesley on analyzing framing strategies of twitter texts using ML, NLP and causal inference analysis.


In 2019, I got MS degree from Operation Research, Management Science & Engineering Department at Stanford. During my graduate study, It was my honor to be advised by Professor Chiara Sabatti on statistical testing and graph-based methods for unsupervised learning.


In 2017, I received BS degree from Civil Engineering, Tongji University and worked on traffic system optmization and traffic emission problems under the guidance of Professor Zhiguo Yan.


Here is my academic CV for research experiencess, and industrial resume for work experiences. I look forward to hearing from you through Email : jim.li.stanford2019@gmail.com

PROFESSIONAL EXPERIENCE

Placeholder image

Amazon 126 Lab

Software Engineer -

  • Online Learning for Type-ahead: according to user typing history on keyboards and current popular keyword trending, to predict and fill in the rest of word searching using NLP technologies.
c++ Python OOD ML NLP system design
Placeholder image

Cadence Design Systems

Software Engineer (R&D) -

  • Balanced Graph Partition: Implemented graph partition algorithms from literature (Kernigan & Lin, spectral clustering, Modularity partition) to improve parallelism of distributed computing, decreased customer design imbalance by 70%, reduced runtime from 20hrs to 11hrs and won key contributor award.
  • C++ Datastream Enhancement: support 64bit infrastructure and implemented efficient datastream compression, reduced 12% memory cost.
c++ Python VLSI graph partition infrastructure
Placeholder image

Stanford CS Department

Teaching Assistant -

  • CS229T Statistical Learning Theory.

RESEARCH WORK

Publications

1. Chao Guo, Zhiyuan Li, Hehua Zhu, Li Zhao, Zhiguo Yan, Single-channel blowing-in longitudinal ventilation method and its application in the road tunnel, Tunnelling and Underground Space Technology (IF=4.45) Volume 108, 2021. [PDF]
2. Willow Wu, Zhiyuan Li, Chuck Eesley, Beyond Word Matching: Measure Framing Strategy Using Machine Learning Methods (WIP)
3. Zhiyuan Li, Pei Cao, and Junchao Gui. “Research on Micro-geometry of Sea Sands Using Scanning Microscope and Particle Analyzer.“ GeoShanghai International Conference. Springer, Singapore, 2018.
4. Jiang, Mingjing, Zhiyuan Li, Hepeng Huang, and Jun Liu. “Experimental study on microstructure and mechanical properties of seabed soft soil from South China Sea”. Chinese Journal of Geotechnical Engineering, Vol. 39 (2), 2017.


Software Patents

1. Yan, Zhiguo, Zhiyuan Li, Li Zhao. “Single Channel Air Supply Type Longitudinal Ventilation Method”. Software Copyright Number: 2016SR133527
2. Yan, Zhiguo, Zhiyuan Li. “Evaluation System for Quality of Road Tunnel Lining”. Software Copyright Number: 2016SR121826

RESEARCH PROJECTS

Analyzing Twitter Framing Strategy

with Prof. Charles Eesley, Stanford MS&E, 1/2019-present

kNN Graph for Feature Selection

with Prof. Chiara Sabatti, Stanford Statistics Dept. 05/2018-09/2018

Benchmarking for Statistical Testing

with Prof. Chiara Sabatti, Stanford Statistics Dept. 09/2018-12/2018

Traffic Ventilation Optimization

with Prof. Zhiguo Yan, Tongji CEE Dept. 04/2015-10/2016

Real-time Human Voice Detection

Course project with Prof. Ashish Goel, Stanford MS&E Dept.

Seismic Loss Prediction

Course project with Prof. Jack Baker, Stanford CEE Dept.