A data scientist with over nine years of commercial experience in providing advanced analytic solutions and thirteen years of experience in data-related roles. I have hands-on experience in manipulating various sources of data at any scale to deliver actionable insights. I also have an in-depth understanding of machine learning, especially deep learning theory, and real-world experience applying it to solve problems.
Responsible for the development of life cycle health management technology for EV cars, which impacts 10 million in-use EV cars in China. Using both data-driven models and mechanism models, I have developed battery and motor safety diagnoses and state of health estimation algorithms. I manage two teams comprising 40 colleagues; one team is responsible for algorithm design, and the other is responsible for system development
In charge of the entire back-end data platform supporting the fintech business, and manage four teams consisting of 24 colleagues grouped as follows:
1. Data acquisition team for internet data crawler and API development.
2. The data engineering team for data cleaning, governance, and advanced analytics.
3. Risk modeling team for credit risk and management modeling.
4. Machine learning research team for fundamental algorithm application research.
Led a small team of data engineers to utilize big data and AI technologies that support the various aspects of the company’s business.
Achievements:
Created and maintained a near real-time dashboard to visualize the performance of the entire company’s global business, based on SSIS and Power BI.
Coordinated with the mobile internet usage supplier to optimize the online game advertisement delivery strategy, resulting in a user click rate increase of over 10 times. Used user portraits and my unique decision forest classifier.
Developed a knowledge graph-based data matching system that helped the NSW government recover fraudulent first home buyer’s grants of more than 70 million dollars.
Provided factual evidence for operations such as unlicensed car dealers and locating missing persons in NSW, which led to receiving the Secretary’s Award of Excellence in Achievement from NSW DFSI.
Additionally, developed a data sharing system that significantly reduced the time for investigators to obtain proper intelligence from several days to several minutes. By the director’s estimation, this system also reduced claim referral time by 90%.
The system achieved 99.6% prediction accuracy for on-time lodgers, and over 70% prediction accuracy for non-lodgers.
With the implementation of this system, ATO has significantly increased its operational efficiency by reducing the number of unnecessary contacts with lodgers, resulting in a 50% decrease in operational costs.
I have also been involved in many other projects, such as Microsoft Australia, the Department of Finance, and the University of Technology, where I utilised advanced techniques like transfer learning and coupled analysis.
Worked as a data engineer, developing several data integration and visualisation utilities for major telecommunication companies in China. Helped these companies improve their risk reporting system by using the Microsoft Business Intelligence pack. At Suntel, a golden partner of Microsoft, delivering customized BI solutions to the clients. Gained experience in the full lifecycle of BI projects, from consulting to designing and deployment, and this kickstarted my career in the data industry.