- Automated network switch design and optimization process, saving manual efforts from days to minutes as well as lowering latency, power assumption and chip area by 5-20%- Formulated the switch design task to a multi-objective constrained optimization problem and abstracted the service input to a directed acyclic graphs (DAG)- Deployed greedy, backtrack, genetic and heuristic algorithms to find optimized solution as well as improve the efficiency of searching process- Collaborated closely with colleagues, working together on Git, reading cutting-edge papers and sharing ideas regularly

- Acted as a full-stack engineer to design and develop a MVC-based website for NSO- Encapsulated the algorithm, created MySQL database and managed the flow of data between frontend and backend- Applied React to generate user-friendly GUI and provide visualized chip design result to the client- Communicated with the clients to discuss about enhancements, collect feedbacks and introduce new features

- Applied Python Scrapy to develop a web crawler that can collect job postings from nearly 100 target companies in batch and store records in MySQL database- Operated and maintained the internal website that shows job postings with several filters for posted date, job title, functionalities and skills needed

- Developed data pipeline to automatically send emails attached with air quality and nearby facilities' report to users when receiving complaints and forward the complaints to the main regulator- Applied Shell and Python scripts to compute pollutant concentration at the user’s location based on NOAA’s HYSPLIT dispersion model

- Give lectures and write tutorials of statistics theory (distributions, simulation, testing, regression) and R programming (dplyr, ggplot2, stats)

- Applied Python crawler to collect more than 1 million emission data and cleaned data;- Summarized Chinese emission data collection framework through doing policy review, and designed an environmental performance assessment model including 103 criteria;- Assessed more than 22,000 enterprises’ environmental performance from 2012 to 2015 to identify the key area and industry for emission control in China.

- Revised and managed green finance dataset and created a data collection and review flow chart to standardize the dataset management process.- Tracked latest green finance news in China and collaborated with colleagues in DC and Chicago offices to write global green finance weekly reports.

- Analyzed wastewater treatment system in Xinji Industrial Park based on Agent-Based Model (ABM);- Developed a wastewater treatment cost estimation model using SPSS, calculated treatment cost of centralized and decentralized methods and found that a centralized method could save 18.7% of cost;- Applied MATLAB to find an optimal pretreatment standard utilizing a centralized method, which could further save 36.2% of total treatment costs while meeting final effluent standards.

- Conducted literature research on financial mechanisms and services of more than 20 green banks around the world and finished a green bank policy memo to senior management team.