1. Designed and completed projects to support new business forms and promotion activities in collaboration with PD, PM and TE, e.g., offered customers options to purchase fund products instead of withdraw when their pension insurance due and boosted yearly AUM by 1.1 billion USD, let customers purchase 3 fund products with coupons in one step and have attracted 2 million new customers, provided alternative asset management group product by custody transfer and increased yearly AUM by 0.5 billion USD, supported tax-deferred commercial pension insurance business development, integrated test and load test with CSDC, CIB and securities institutions, using Java, Spring, SQL, RPC framework, distributed message framework, distributed cache framework, distributed task scheduling framework, distributed transaction framework, monitoring framework, etc.

2. Supported market regulations related projects, e.g., forbade purchase from online verification failed customers and guided them to update personal information, developed and tested system modification due to isolation between main site and fund site.

3. Propelled system optimization with abstraction, coordinated upstream and downstream systems to migrate historical links to standard links, provided general solution for various promotion activities demand, greatly improved development efficiency by 60% and decreased modification risks.

1. Accomplished spark log colorization based on log types using Angular and Typescript to help users efficiently check log content.

2. Improved log searching by highlighting matched content and adding Enter key response to iterate through all matched content based on Monaco Editor.

3. Added load more function for spark log, helping users get updated and historical log without refreshing the whole page and greatly decrease waiting time, and provided log filter function to only show error and warning content when check box is selected.

1. Accomplished projects of pulmonary nodules diagnosis, gastric cancer phases classification based on machine learning and deep learning, using python, scikit-learn, pytorch.

2. Devised a structure combined with feature selection, model building, model evaluation based on machine learning, radiomics and python, thus dramatically reduced time of processing new project by 60%.